CN-122000895-A - Power system transient stability control method and device based on active low-voltage ride through of wind turbine generator

CN122000895ACN 122000895 ACN122000895 ACN 122000895ACN-122000895-A

Abstract

The invention discloses a transient stability control method and device for a power system based on active low-voltage ride through of a wind turbine. The method takes a near-end strategy optimization algorithm as a core to construct a reinforcement learning model, wherein a grid-connected power system with a fan is taken as an environment, multi-time characteristic electric quantity of a node model is extracted to construct an observation state, and the end time of active low-voltage ride through is taken as a control action. And through training, the near-end strategy optimization algorithm network learns to decide the optimal low-pass ending time, and generates an optimal control strategy for enabling the transient state of the system to be quickly restored to be safe. The method can rapidly and accurately give out transient stability control measures when the system is disturbed.

Inventors

SUN ZHENGLONG
WANG SIXUAN
ZHANG RUI
LI ZEWEI
YANG HAO
CAI GUOWEI

Assignees

东北电力大学

Dates

Publication Date: 20260508
Application Date: 20251229

Claims (9)

1. The power system transient stability control device based on the active low-voltage ride through of the wind turbine generator is characterized by comprising an active low-voltage ride through module, an active reactive power optimization control module, an initialization module, an interaction module, a data processing module, a decision module and a learning module.
2. The wind turbine active low voltage ride through-based power system transient stability control device of claim 1, wherein the device is characterized by: The active low-voltage ride through module is used for controlling active switching of a crowbar circuit of the wind turbine generator and active locking of a rotor converter; the active and reactive power optimization control module is used for optimizing the low-voltage ride through process, actively controlling active and reactive power of the fan, controlling active output of the system and adjusting the system voltage level influenced by disturbance; the initialization module is used for configuring network parameters of a reinforcement learning network based on a proximity optimization strategy algorithm, setting the maximum interaction times of each cycle among interactions and the number of cycles to be trained, and reading a preset low-voltage ride-through ending instruction set and a characteristic electrical quantity set of the power system; the interaction module is used for reading the characteristic electrical quantity data once after the power system runs an interaction interval step, and transmitting the instruction to the power system if the reinforcement learning network has instruction output of ending the low voltage ride through; the data processing module is used for dividing the characteristic electric quantity read by the interaction module into three types, namely decision measure electric quantity, control effect electric quantity and safety constraint electric quantity; The evaluation module is used for obtaining a reward value by using a reward function according to the voltage amplitude value and the phase angle value of each node of the system; the decision module is used for transmitting the decision measure electric quantity as input data of the reinforcement learning network to the reinforcement learning network, and transmitting an active low-voltage ride-through ending control instruction of the wind turbine generator as output of the reinforcement learning network, so that the wind turbine generator can give out corresponding active low-voltage ride-through measures when a power system is disturbed, and the transient stability of the system is restored to a safe range; The learning module is used for judging the transient stability recovery effect of the system according to the electric quantity of the control effect and judging whether to trigger the safety constraint of the power system according to the electric quantity of the safety constraint, and based on the judgment, the network parameters of the reinforcement learning network are updated by combining the rewarding value obtained by the judging module.
3. The electric power system transient stability control device based on the wind turbine active low voltage ride through of claim 2 is characterized in that the reinforcement learning network based on the proximity optimization strategy algorithm comprises two neural networks, namely a strategy neural network and a value neural network, wherein the input of the strategy neural network is real-time node voltage amplitude and phase angle, the output of the strategy neural network is a wind turbine active low voltage ride through ending control instruction, the input of the value neural network is a system stability index sBTTC value judged after the strategy is executed, the output of the value neural network is a neural network weight used for updating the strategy neural network and the value neural network, the characteristic electric quantity comprises node voltage amplitude and phase angle of a plurality of time points in the latest interaction interval, synchronous machine power angle difference, active power, reactive power of each machine set, system bus voltage threshold and low voltage ride through time threshold, the network parameters of the reinforcement learning network based on the proximity optimization strategy algorithm comprise learning rate, batch size, gradient clipping size and discount factor size, and the wind turbine active low voltage ride through ending control instruction set is constructed by a crowbar circuit disconnection instruction and a rotor reconnection instruction.
4. The power system transient stability control device based on wind turbine active low voltage ride through of claim 3, wherein the interaction interval is the time of how often the reinforcement learning network interacts with the power system, and each interaction interval is set to be 1 second.
5. The transient stability control device of the power system based on the active low-voltage ride through of the wind turbine generator set, which is characterized in that the decision measure electric quantity comprises a real-time node voltage amplitude value and a phase angle, the decision measure electric quantity is used as an input value of a reinforcement learning network, the control effect electric quantity comprises a synchronous machine power angle difference, the transient stability control instruction control effect given by the reinforcement learning network in the last interaction interval is judged according to the power angle difference reduction degree, and the safety constraint electric quantity, namely a low-pass time threshold, comprises the active power, the reactive power, a system bus voltage threshold and a low-voltage ride through time threshold of each wind turbine generator set.
6. The power system transient stability control device based on wind turbine active low voltage ride through of claim 2, wherein the input of the reinforcement learning network in the decision module is decision measure electric quantity, and the output is wind turbine active low voltage ride through end control instruction, and the active low voltage ride through end instruction comprises active disconnection of a crowbar circuit and active reconnection of a rotor converter.
7. The wind turbine active low voltage ride through-based power system transient stability control device of claim 2, wherein the reward function is set as follows: If sBTTC index value >0.6, setting the bonus function to BTTC i ×1000; if sBTTC index values are within the range of [0.2,0.6], the prize is set as a linear function: (BTTC i -0.2)×1000; If sBTTC index value <0.2, the bonus function is set to BTTC i X (-1000).
8. A method of using a wind turbine active low voltage ride through based power system transient stability control device according to any of claims 1 to 7, comprising: Firstly, building a power system model containing an active low-voltage ride-through wind turbine, wherein the active low-voltage ride-through capability of the wind turbine is realized through hardware control and software control, the hardware control content is an active switching crowbar circuit and an active locking rotor converter, and the software control content is the optimization of the active reactive current to low-voltage ride-through process; Extracting characteristic electrical quantities of a plurality of historical moments of a power system node model as state values in reinforcement learning to construct observation data, wherein the characteristic electrical quantities of the plurality of historical moments of the power system node model comprise node voltage amplitude values and phase angles at a plurality of time points in an interaction interval, synchronous machine power angle differences, unit active power, reactive power, system bus voltage thresholds and low voltage crossing time thresholds; The method comprises the steps of constructing a reinforcement learning network based on a near-end strategy optimization algorithm, wherein the reinforcement learning network based on the near-end strategy optimization algorithm comprises a strategy neural network and a value neural network, the input of the strategy neural network is real-time node voltage amplitude and phase angle, and the output of the strategy neural network is a system transient stability control instruction for the end of active low-voltage ride through of a wind turbine generator; The characteristic electric quantity is analyzed and classified, the characteristic electric quantity is divided into decision measure electric quantity, control effect electric quantity comprises power angle difference reduction and safety constraint electric quantity, the decision measure electric quantity comprises real-time node voltage amplitude and phase angle, the decision measure electric quantity is used as an input value of a reinforcement learning network, the control effect electric quantity comprises synchronous machine power angle difference, transient stability control instruction control effect given by the reinforcement learning network in an interaction interval is judged according to the power angle difference reduction degree, and the safety constraint electric quantity comprises low-pass time threshold values comprising active power, reactive power, system bus voltage threshold values and low-voltage pass time threshold values of all units; step five, utilizing the reinforcement learning network based on the adjacent optimization strategy algorithm to perform optimization training on the characteristic electric quantity data to generate a strategy model; And step six, extracting the real-time decision measure electric quantity of the electric power system, taking the electric quantity as the input of a strategy model, outputting an active low-voltage ride through ending control instruction of the wind turbine generator, and sending the output instruction to the electric power system for execution.
9. The method for controlling transient stability of a power system based on active low-voltage ride through of a wind turbine generator set according to claim 8 is characterized in that in the fifth step, the strategy model is a strategy neural network structure and parameters in a training-completed reinforcement learning network based on a proximity optimization strategy algorithm, an input layer is decision-making measure electric quantity, and an output is a wind turbine generator set active low-voltage ride through ending control instruction.

Description

Power system transient stability control method and device based on active low-voltage ride through of wind turbine generator Technical Field The invention relates to the technical field of transient stability control of power systems, in particular to a power system transient stability control system and method based on active low-voltage ride through of a wind turbine generator. Background In recent years, china actively and steadily advances the carbon to reach the peak carbon neutralization, and steadily advances the green low-carbon transformation of energy, so that the development of new energy and clean energy is promoted at a more prominent position. In a traditional power system mainly based on centralized thermal power and hydro-electric power, the transient stability of a power grid mainly depends on the rotor inertia and the regulating capability of a synchronous generator. However, the dynamic characteristics of the power system, in particular the introduction of the doubly-fed induction generator and the full-power wind turbine, are changed by the connection of the wind turbine, so that the inertia of the system is obviously reduced, and a new threat is formed to the transient stability of the system. Therefore, the method has important significance in transient stability control of the wind turbine generator system. The low voltage ride through capability of the wind turbine generator can change the power balance of the system and further serve as a transient stability control means, but the traditional low voltage ride through is affected by the voltage drop degree and the rotor current out-of-limit degree, the response speed is low, the flexibility is poor, so that the low voltage ride through of the wind turbine generator is actively controlled according to the transient instability condition of the system, the transient stability control of the power system is carried out by utilizing the self-coping fault protection behavior of the wind turbine generator, and more control means can be provided for the transient stability control of the power system of the grid connection of the wind turbine generator. In recent years, in practical engineering cases at home and abroad, many wind farms successfully solve the problem of system instability caused by off-grid wind turbine generators by improving a low-voltage ride through technology. Therefore, the transient stability of low voltage ride through of the fan-containing system is studied in depth, which is not only helpful for improving the safety and stability of the existing power grid, but also lays a solid foundation for promoting the further development of new energy power generation. The research is crucial in practical engineering application, and directly influences the reliability and the sustainability of the future large-scale wind power grid connection. Content of the application Aiming at the problems, the power system transient stability control device based on active low voltage ride through of the wind turbine, provided by the invention, comprises: The system comprises an active low voltage ride through module, an active reactive power optimization control module, an initialization module, an interaction module, a data processing module, a decision module and a learning module. Further, the active low-voltage ride through module is used for controlling active switching of a crowbar circuit of the wind turbine generator and active locking of a rotor converter; the active and reactive power optimization control module is used for optimizing the low-voltage ride through process, actively controlling active and reactive power of the fan, controlling active output of the system and adjusting the system voltage level influenced by disturbance; the initialization module is used for configuring network parameters of a reinforcement learning network based on a proximity optimization strategy algorithm, setting the maximum interaction times of each cycle among interactions and the number of cycles to be trained, and reading a preset low-voltage ride-through ending instruction set and a characteristic electrical quantity set of the power system; the interaction module is used for reading the characteristic electrical quantity data once after the power system runs an interaction interval step, and transmitting the instruction to the power system if the reinforcement learning network has instruction output of ending the low voltage ride through; the data processing module is used for dividing the characteristic electric quantity read by the interaction module into three types, namely decision measure electric quantity, control effect electric quantity and safety constraint electric quantity; The evaluation module is used for obtaining a reward value by using a reward function according to the voltage amplitude value and the phase angle value of each node of the system; the decision module is used for transmitting the decision measure electric