CN-122026495-A - DPWM-based network-structured converter overload capacity improving method and system

CN122026495ACN 122026495 ACN122026495 ACN 122026495ACN-122026495-A

Abstract

The invention belongs to the technical field of power electronic intelligent control, and provides a DPWM-based method and a DPWM-based system for improving overload capacity of a grid-structured converter, which comprise the steps of obtaining d/q axis current instructions of the converter, synthesizing current instruction included angles, constructing a data set of the current instruction included angles, clamping offset angles and junction temperatures, and establishing direct nonlinear mapping from the current instruction included angles to optimal clamping offset angles through an artificial neural network model; under overload condition, the optimal clamping offset angle is output by the model quickly by using the real-time current command included angle, the three-phase voltage command extremum is extracted and transformed, the optimal clamping offset angle is injected into the electric angle to make sector selection, the zero sequence component injection voltage command is generated by combining the results, the clamping mode is switched dynamically to enable the clamping interval to be aligned to the current peak area, and the zero switching action of the peak section of the insulated gate bipolar transistor is realized to inhibit junction temperature. The invention simplifies the calculation, improves the response speed, precisely suppresses the junction temperature and obviously improves the overload capacity of the converter.

Inventors

LIU ZHIJIE
TAO YE
LI KEJUN
Xiang Yanming
Su Baihe
YING HAIZHOU
KUANG YUXIANG
SUN YUANYUAN
SUN KAIQI

Assignees

山东大学

Dates

Publication Date: 20260512
Application Date: 20260414

Claims (10)

1. The DPWM-based network-structured converter overload capacity improving method is characterized by comprising the following steps of: acquiring a d-axis current instruction and a q-axis current instruction of the grid-formed converter, and combining the d-axis current instruction and the q-axis current instruction into a current instruction included angle; Constructing a data set of a current instruction included angle, a clamping offset angle and junction temperature based on multi-physical field thermal simulation, and further constructing a direct nonlinear mapping from the current instruction included angle to an optimal clamping offset angle through an artificial neural network model; under the overload working condition of the converter, the optimal clamping offset angle is rapidly output through an artificial neural network model according to the real-time current instruction included angle; And simultaneously carrying out extremum extraction on the three-phase voltage command, carrying out coordinate transformation on the three-phase voltage command to obtain a fundamental wave phase angle, superposing an optimal clamping offset angle and the fundamental wave phase angle to carry out sector selection, and generating a zero sequence component injection voltage command by combining the sector selection and extremum extraction result so as to dynamically switch a clamping mode, so that a clamping interval is aligned to a current peak area, and further, an insulated gate bipolar transistor acts in a zero switching mode at the current peak section to inhibit junction temperature, thereby improving the overload capacity of the grid-structured converter.
2. The method for improving overload capacity of a grid-structured converter based on DPWM according to claim 1, wherein the formula of the combined current instruction included angle of the d-axis current instruction and the q-axis current instruction is: ; In the formula, For the d-axis current command, For the q-axis current command, Is the current command included angle.
3. The method for improving overload capacity of a grid-structured converter based on DPWM according to claim 2, wherein the data set is expressed as: ; The mapping relationship between the junction temperature and the input variable can be expressed as: ; In the formula, In order to clamp the angle of offset, In order to achieve the junction temperature, In the case of a data set, As the minimum value of the current angle sweep, At the maximum value of the current angle sweep, In order to shift the minimum value of the phase angle, Is the maximum value of the phase shift angle.
4. A method for improving overload capacity of a grid-structured converter based on DPWM according to claim 3, wherein said direct nonlinear mapping is as follows: ; In the formula, For an optimal clamping offset angle, An artificial neural network model.
5. The method for improving overload capacity of a grid-structured converter based on DPWM according to claim 1, wherein aligning the clamping interval to the current peak area includes extremum selection, coordinate transformation, sector selection and zero sequence component injection.
6. The method for improving overload capacity of a grid-structured converter based on DPWM according to claim 5, wherein said extremum is selected as: ; In the formula, Is a three-phase reference voltage input.
7. The method for improving overload capacity of a grid-structured converter based on DPWM according to claim 6, wherein the coordinate transformation obtains a fundamental phase angle by the formula: ; In the formula, For the phase angle of the fundamental wave, As a component of the voltage on the alpha axis, Is the beta-axis voltage component.
8. The method for improving overload capacity of a grid-structured converter based on DPWM as set forth in claim 7, wherein said sector selection formula is: ; In the formula, In order to control the angle of the angle, Is a logical sector.
9. The method for improving overload capacity of a grid-structured converter based on DPWM according to claim 8, wherein the formula of zero sequence component injection is: ; In the formula, In order to finally modulate the wave, Is the original unclamped modulated wave, Is the zero sequence component injected.
10. A DPWM-based grid-tied converter overload capability promotion system for implementing a DPWM-based grid-tied converter overload capability promotion method according to any one of claims 1 to 9, characterized by comprising: the characteristic extraction module is used for obtaining a d-axis current instruction and a q-axis current instruction of the grid-built current transformer and synthesizing to obtain a current instruction included angle; the electrothermal mapping module is used for constructing a data set of a current instruction included angle, a clamping offset angle and junction temperature based on the multi-physical-field thermal simulation data, and establishing direct nonlinear mapping from the current instruction included angle to an optimal clamping offset angle through an artificial neural network model; The real-time decision module is used for outputting an optimal clamping offset angle through the artificial neural network model according to the real-time current instruction included angle under the overload working condition of the converter; The dynamic modulation execution module is used for injecting the optimal clamping offset angle into a modulation signal generation path, dynamically switching a clamping mode through sector logic, enabling a clamping interval to be aligned to a current peak area, and realizing zero switching action of the insulated gate bipolar transistor at the current peak section so as to inhibit junction temperature.

Description

DPWM-based network-structured converter overload capacity improving method and system Technical Field The invention belongs to the technical field of power electronic intelligent control, and provides a DPWM-based method and system for improving overload capacity of a grid-built converter. Background With the rapid development of a high-proportion new energy power system, a grid-structured converter has become core equipment for the stable control of new energy grid connection, micro-grids and distribution networks by virtue of active voltage and frequency supporting capability. When voltage drop, short circuit fault or severe load fluctuation occurs in the power grid, the grid-structured converter needs short-time overload operation to ensure the stability of the power grid, the output current of the converter is rapidly increased at the moment, the electrothermal stress of a power device is obviously increased, and the device is extremely easy to lose efficacy due to junction temperature overrun, so that the thermal safety of the device and the overload tolerance capability of the converter under the overload working condition are improved, and the grid-structured converter becomes a key technical requirement for engineering application. Discontinuous Pulse Width Modulation (DPWM) is widely used for inhibiting junction temperature of a power device because switching actions and switching loss can be reduced, and is characterized in that a DPWM clamping offset angle is reasonably set, so that a clamping interval is matched with a loss hot spot to realize optimal thermal control. The existing DPWM clamp angle optimization method is mainly based on a Switching Loss Function (SLF) to carry out analytic calculation, a power factor angle is solved by collecting voltage and current signals, and then an optimal clamp angle is determined based on a loss model. However, the method has the obvious defects that firstly, the traditional analytical model only considers the switching loss, the mapping relation between the total loss of the device and junction temperature cannot be completely represented, the optimal clamping angle is deviated from the actual minimum point, secondly, the control process needs to introduce multidimensional variables such as a voltage phase angle, a current phase angle, a power factor angle and the like, the calculation link is complex, the real-time operation amount is large, obvious hysteresis exists in the transient process of millisecond-level rapid increase of the overload current, the quick response is difficult, thirdly, the traditional method relies on multi-physical-quantity sampling and complex model calculation, the complexity of software and hardware of the system is high, the parameter drift is easily influenced under the network-structured control architecture, and the thermal control precision and the robustness are difficult to meet the requirement of overload working conditions. Disclosure of Invention In order to solve the technical problems, the invention provides a DPWM-based method and a DPWM-based system for improving overload capacity of a grid-structured converter, which can simplify calculation, improve response speed, accurately inhibit junction temperature and remarkably improve overload capacity and operation reliability of the converter. The technical scheme of the invention comprises the following steps: And acquiring a d-axis current instruction and a q-axis current instruction of the grid-formed converter, and combining the d-axis current instruction and the q-axis current instruction into a current instruction included angle. And constructing a data set of the current instruction included angle, the clamping offset angle and the junction temperature based on multi-physical field thermal simulation, and further constructing direct nonlinear mapping from the current instruction included angle to the optimal clamping offset angle through an artificial neural network model. And under the overload working condition of the converter, the optimal clamping offset angle is rapidly output through an artificial neural network model according to the real-time current instruction included angle. And simultaneously carrying out extremum extraction on the three-phase voltage command, carrying out coordinate transformation on the three-phase voltage command to obtain a fundamental wave phase angle, superposing an optimal clamping offset angle and the fundamental wave phase angle to carry out sector selection, and generating a zero sequence component injection voltage command by combining the sector selection and extremum extraction result so as to dynamically switch a clamping mode, so that a clamping interval is aligned to a current peak area, and further, an insulated gate bipolar transistor acts in a zero switching mode at the current peak section to inhibit junction temperature, thereby improving the overload capacity of the grid-structured converter. Further, the form