A transformer control assembly with self-diagnosis function and a fault early warning method thereof

Abstract

The application discloses a transformer control assembly with self-diagnosis function and a fault early warning method thereof. The assembly comprises a main power conversion module, a high-frequency pulse injection branch, a phase-locked timing controller, a echo capture and digitization unit and a processing unit. The phase-locked timing controller monitors the dead time of the power loop and triggers the detection pulse injection during the period. The core of the application is to use the high impedance condition formed by the full-off of the power device in the dead time, so that the reflection coefficient of the injection point tends to 1, and the physical level gain amplification of the reflected voltage echo is realized; by setting the pulse round-trip propagation time to be less than, it is ensured that the echo falls into the high impedance window to avoid power flow interference. The processing unit establishes the equivalent impedance spectrum of the winding according to the captured echo and identifies the resonance peak offset to output a graded early warning signal. The application breaks through the technical prejudice of the dead time mute protection, and significantly improves the online monitoring sensitivity of the deformation of the internal winding of the transformer.

Description

Technical Field

[0001] This invention relates to the intersection of online monitoring of power transformers and power electronic control technology, specifically to a transformer self-diagnostic control component and early warning method that utilizes the dead time of the control loop for impedance spectrum detection. Background Technology

[0002] As a critical piece of equipment in the power system, the operating status of power transformers directly affects the safety and stability of the power grid. Current technologies for monitoring transformer operating status typically rely on external sensing methods, such as oil chromatography, partial discharge detection, vibration monitoring, and ultrasonic testing. While these methods can reflect the internal condition of the transformer to some extent, they generally suffer from the following shortcomings: First, they require additional sensor installation, which is complex and costly; second, the detection results are easily affected by the electromagnetic environment at the site, resulting in poor stability; third, most methods are indirect or delayed detection methods, with limited sensitivity to minor winding deformations or early insulation degradation, making it difficult to provide timely early warnings.

[0003] Meanwhile, in the field of power electronics control, transformer control components (such as power conversion circuits used in on-load tap changers or related auxiliary systems) typically employ pulse width modulation (PWM) driving. In this type of control strategy, a dead time is generally set to prevent shoot-through short circuits in the upper and lower bridge arm power devices. Existing technologies generally treat this dead time as an "ineffective period" solely for protection, during which no signal utilization or information acquisition occurs, and its potential value is not explored.

[0004] In addition, for winding condition identification, some technologies have attempted to determine structural parameters by injecting signals and analyzing reflection characteristics or impedance spectrum changes. However, such methods mostly rely on independent testing devices or dedicated detection circuits, which are difficult to integrate deeply with existing control systems. Furthermore, under high-power operating conditions, the signal is easily submerged by the main power flow, resulting in a low signal-to-noise ratio and affecting detection accuracy.

[0005] Therefore, how to achieve highly sensitive online detection of the internal state of a transformer without adding complex external detection equipment, and how to improve the signal-to-noise ratio of the detection signal, thereby enabling effective early warning of early faults such as winding deformation, has become a technical problem that urgently needs to be solved in this field. Summary of the Invention

[0006] The technical problem to be solved by this invention is: how to overcome the problems of existing transformer monitoring methods relying on external sensors, being susceptible to power flow interference, and having low detection sensitivity, and to provide a control component and method that can achieve high signal-to-noise ratio detection by utilizing the inherent gap of the control loop.

[0007] The above-mentioned technical objective of the present invention is achieved through the following technical solution: a transformer control component with self-diagnostic function, comprising:

[0008] The main power conversion module has a preset dead time. ; A high-frequency pulse injection branch is connected to the output terminal of the main power conversion module; A phase-locked timing controller is communicatively connected to both the main power conversion module and the high-frequency pulse injection branch. It monitors the switching state of the main power conversion module and, within the dead time... The internal trigger injects a high-frequency detection pulse into the transformer side via the high-frequency pulse injection branch. The echo capture and digitization unit is used to capture the reflected voltage echo from the transformer side under the high impedance condition formed when the main power conversion module is turned off. ;as well as The processing unit is connected to the echo acquisition and digitization unit; Wherein, the round-trip propagation time of the high-frequency detection pulse on the transformer side Less than the dead time .

[0009] Furthermore, the component also includes a wideband capacitive coupler, through which the power flow generated by the main power conversion module and the detection pulse generated by the high-frequency pulse injection branch are combined and output to the transformer to be monitored.

[0010] Furthermore, the high-frequency pulse injection branch is used in the dead time Internally, by utilizing the high impedance state presented by the main power conversion module, the reflection coefficient at the injection point is reduced. To meet the high signal-to-noise ratio reflection condition, thereby enhancing the reflected voltage echo. The signal strength.

[0011] Furthermore, the processing unit is used to process the reflected voltage echo. Calculate the equivalent input impedance spectrum of a transformer winding and from the equivalent input impedance spectrum Extracting the real-time detection resonant frequency .

[0012] Furthermore, the processing unit is also connected to the phase-locked timing controller for processing based on the captured round-trip time. Feedback is used to adjust the dead time control parameters of the phase-locked timing controller.

[0013] The present invention also provides a transformer fault early warning method using the above-mentioned components, comprising the following steps: Step S1: Monitor the switching state of the main power conversion module in real time and identify the dead time window when the power device enters the fully off high-impedance state. ; Step S2: In the dead time window A high-frequency probe pulse is injected internally, and the enhanced reflected voltage echo under the high-resistivity state is acquired. ; Step S3: Based on the reflected voltage echo Establish the equivalent input impedance spectrum of the transformer winding and from the equivalent input impedance spectrum Extracting the real-time detection resonant frequency ; Step S4: Calculate the real-time detection resonant frequency Relative to the reference resonant frequency resonance peak shift ; Step S5: Shift the resonance peak amount The signal is compared with a preset threshold, and an early warning signal for the deformation state of the transformer's internal windings is output based on the comparison result.

[0014] Furthermore, a baseline establishment step is included before step S1: A reference scan is performed when the transformer is in a healthy state to obtain the reference scan impedance spectrum of the transformer winding. The corresponding peak frequency is recorded as the reference resonant frequency. .

[0015] Furthermore, in step S2, the high-frequency detection pulse is precisely synchronized and controlled within the dead time using a phase-locked loop (PLL). Injection is triggered 10%-30% of the time after the start.

[0016] Furthermore, in step S5, based on the resonance peak shift... The corresponding physical displacement level triggers a graded early warning: when At that time, a level 2 alert will be output; when At that time, a Level 1 severe warning will be issued; in, and This is the preset frequency offset threshold.

[0017] Compared with the prior art, the present invention has the following beneficial effects: 1. This invention injects a high-frequency detection pulse during the dead time of the main power conversion module, utilizing the high impedance physical characteristics of the power device in the off state to reduce the reflection coefficient at the detection signal injection point. Approaching 1. This invention achieves reflected voltage echo through this physical-level reflection enhancement mechanism. The significant improvement in signal-to-noise ratio solves the problem that traditional online monitoring signals are easily overwhelmed by high-current power noise.

[0018] 2. This invention sets the round-trip propagation time of the detection pulse on the transformer side. Less than dead time This invention achieves complete decoupling of the detection signal and the main power control signal in the time domain. Through this time-scale constraint, the invention ensures that the capture process of the reflected echo occurs entirely during the off-state period of the main circuit power devices, eliminating electrical interference from the power flow to the diagnostic signal.

[0019] 3. This invention compares the impedance spectra of a reference scanning system. With real-time detection impedance spectrum The difference was used to extract the resonance peak shift. As a criterion, this invention achieves non-destructive identification of the mechanical displacement and deformation state of the transformer's internal windings by monitoring the frequency domain shift of equivalent electrical parameters, thereby improving the accuracy of early warning of initial faults.

[0020] 4. This invention uses a processing unit to determine the round-trip propagation time based on the captured data. Feedback adjustment of the phase-locked loop timing controller parameters enables adaptive matching between diagnostic logic and main circuit operating status, ensuring a stable detection window under different load conditions. Attached Figure Description

[0021] Figure 1 This is a schematic diagram of the overall architecture of a transformer control component with self-diagnostic function provided in an embodiment of the present invention; Figure 2 This is a schematic diagram illustrating the timing relationship between the detection pulse and the power control waveform provided in an embodiment of the present invention; Figure 3 This is a schematic diagram of the high-frequency equivalent circuit model of the transformer-control loop provided in an embodiment of the present invention; Figure 4 This is a schematic diagram showing the impedance spectrum changes of a transformer under different states, as provided in an embodiment of the present invention.

[0022] In the picture: Dead time window; Detect pulse round-trip propagation time; Incident voltage pulse; , reflected voltage echo; Characteristic impedance; Coupling capacitors; Leakage inductance; Winding resistance; Series capacitor between the discs; , , distributed capacitance to ground; Reference resonant frequency; Real-time detection of resonant frequency; , resonance peak offset. Detailed Implementation

[0023] To make the objectives, technical solutions, and advantages of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings.

[0024] In response to the current situation where weak diagnostic signals in transformer online monitoring are easily submerged by the background noise of the main power and the installation of external sensors is complicated, this embodiment provides a transformer control component with self-diagnostic function and its fault early warning method. By reusing the power conversion topology of the control system, characteristic quantities reflecting the mechanical state of the winding are extracted within a specific physical timing window.

[0025] like Figure 1 As shown, the transformer control component with self-diagnostic function in this embodiment includes: a phase-locked loop (PLL) timing controller, a high-frequency pulse injection branch, a main power conversion module, a wideband capacitive coupler, an echo capture and digitization unit, and a processing unit. The main power conversion module, as the core actuator for power conversion, typically employs a bridge inverter or converter circuit. During switching, the integrated power devices generate a preset dead time due to the drive control of the PLL timing controller. The high-frequency pulse injection branch is connected in parallel to the output bus of the main power conversion module via a wideband capacitive coupler. The phase-locked timing controller acquires the carrier state of the main power conversion module in real time, identifies the transient moment when the upper and lower bridge arms are fully turned off, and records the data within the dead time. Within a preset delay after startup, a high-frequency pulse injection branch is triggered to generate an incident voltage pulse. The wideband capacitive coupler physically achieves the electrical superposition of high-frequency probe pulses and low-frequency high-power current, outputting the composite signal to the transformer under monitoring. The echo acquisition and digitization unit monitors the line status in real time through a high-sensitivity sampling network and captures the reflected voltage echoes. After being converted into a discrete numerical sequence, it is transmitted to the processing unit, which executes the impedance identification logic and outputs it to the fault warning and control output.

[0026] The core physical mechanism of this embodiment lies in utilizing dead time. Impedance modulation effect within the window. For example... Figure 2 As shown, in the dead time window Inside, both the upper and lower bridge arm control signals of the main power conversion module are at low levels, and the power devices are in a high-impedance state of complete shutdown. At this time, the impedance observed from the probe signal injection point towards the power source tends to infinity. According to the transmission line reflection coefficient formula, when the load impedance is extremely high, the reflection coefficient... Approaching 1. Compared to the low reflectivity of the main power device in the on-state, the signal reflection efficiency in the high-resistivity state is significantly improved. The round-trip propagation time of the high-frequency probe pulse injected into the high-frequency pulse injection branch within this window is... The closed-loop feedback of the processing unit is strictly controlled within the dead time. Within the range, that is To ensure reflected voltage echo The complete feature envelope falls within the silent window where the main circuit power noise is lowest, eliminating the masking effect of high current power flow on weak diagnostic signals.

[0027] Modeling the winding characteristics of the transformer to be monitored, such as... Figure 3 As shown, the transformer winding under high-frequency excitation can be equivalent to a winding composed of characteristic impedance. Coupling capacitors Leakage inductance Winding resistance series capacitor between discs and distributed capacitance to ground , The distributed parameter network is formed. Among them, the series capacitance between the discs... With distributed capacitance to ground , It is extremely sensitive to the geometric deformation of the winding. When the winding is subjected to radial compression, axial displacement, or local loosening, the relative physical positions between different parts of the winding undergo microscopic changes, directly affecting the equivalent network. or The value has drifted. The processing unit analyzes the reflected voltage echo. The system identification algorithm is executed to calculate the equivalent input impedance spectrum of the transformer winding. .

[0028] like Figure 4 As shown, in specific implementation, the system first performs a reference scan when the transformer to be monitored is in a healthy state to establish a reference scan impedance spectrum. And identify the reference resonant frequency corresponding to its resonant peak. During long-term operation, the processing unit periodically or in real-time extracts the real-time resonant frequency under the current state. Since changes in the winding distributed parameters cause a frequency domain shift in the position of the impedance spectrum poles, the processing unit calculates the resonant peak shift. This allows for a quantitative description of the degree of structural anomalies in the winding. The graded early warning logic set in this embodiment is as follows: when the resonance peak shifts... Less than the preset first threshold When the winding structure is stable, it is determined that the winding structure is stable; when In and When the condition is between these conditions, a microscopic deformation tendency is detected, and a level 2 attention alarm is output; when Exceeding the second threshold When a significant displacement or deformation fault is detected in the winding, a first-level severe warning is output, and emergency protection actions are executed by the fault warning and control output.

[0029] This invention utilizes a phase-locked loop (PLL) timing controller to multiplex the transient characteristics of the power topology. Without adding external sensors or altering the main circuit topology, it leverages the high impedance characteristics of dead time to achieve physical gain of the probe signal, significantly improving the signal-to-noise ratio of the echo. Simultaneously, it utilizes the round-trip propagation time... With dead time The scale matching achieves time-domain decoupling of monitoring signals and energy signals at the physical layer, ensuring that the online detection accuracy reaches the level of static testing, and providing a reliable technical means for the early detection of winding deformation inside transformers.

[0030] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features therein. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A transformer control component with self-diagnostic function, characterized in that, include: The main power conversion module has a preset dead time. ; A high-frequency pulse injection branch is connected to the output terminal of the main power conversion module; A phase-locked timing controller is communicatively connected to both the main power conversion module and the high-frequency pulse injection branch. It monitors the switching state of the main power conversion module and, within the dead time... The internal trigger injects a high-frequency detection pulse into the transformer side via the high-frequency pulse injection branch. The echo capture and digitization unit is used to capture the reflected voltage echo from the transformer side under the high impedance condition formed when the main power conversion module is turned off. ; as well as The processing unit is connected to the echo acquisition and digitization unit; Among them, the round-trip propagation time of the high-frequency detection pulse on the transformer side Less than the dead time .

2. The transformer control component with self-diagnostic function according to claim 1, characterized in that, The component also includes a wideband capacitive coupler, through which the power flow generated by the main power conversion module and the detection pulse generated by the high-frequency pulse injection branch are combined and output to the transformer to be monitored.

3. The transformer control component with self-diagnostic function according to claim 1, characterized in that, The high-frequency pulse injection branch is used during the dead time. Internally, by utilizing the high impedance state presented by the main power conversion module, the reflection coefficient at the injection point is reduced. To meet the high signal-to-noise ratio reflection condition, thereby enhancing the reflected voltage echo. The signal strength.

4. The transformer control component with self-diagnostic function according to claim 1, characterized in that, The processing unit is configured to process the reflected voltage echo. Calculate the equivalent input impedance spectrum of a transformer winding and from the equivalent input impedance spectrum Extracting the real-time detection resonant frequency .

5. The transformer control component with self-diagnostic function according to claim 1, characterized in that, The processing unit is also connected to the phase-locked timing controller, and is used to process the captured round-trip time. Feedback is used to adjust the dead time control parameters of the phase-locked timing controller.

6. A transformer fault early warning method implemented using the components described in any one of claims 1-5, characterized in that, Includes the following steps: Step S1: Monitor the switching state of the main power conversion module in real time and identify the dead time window when the power device enters the fully off high-impedance state. ; Step S2: In the dead time window A high-frequency probe pulse is injected internally, and the enhanced reflected voltage echo under the high-resistivity state is acquired. ; Step S3: Based on the reflected voltage echo Establish the equivalent input impedance spectrum of the transformer winding and from the equivalent input impedance spectrum Extracting the real-time detection resonant frequency ; Step S4: Calculate the real-time detection resonant frequency Relative to the reference resonant frequency resonance peak shift ; Step S5: Shift the resonance peak amount The signal is compared with a preset threshold, and an early warning signal for the deformation state of the transformer's internal windings is output based on the comparison result.

7. The transformer fault early warning method according to claim 6, characterized in that, The benchmark establishment step is included before step S1: A reference scan is performed when the transformer is in a healthy state to obtain the reference scan impedance spectrum of the transformer winding. The corresponding peak frequency is recorded as the reference resonant frequency. .

8. The transformer fault early warning method according to claim 6, characterized in that, In step S2, the high-frequency detection pulse is precisely synchronized and controlled within the dead time using a phase-locked loop (PLL). Injection is triggered 10%-30% of the time after the start.

9. The transformer fault early warning method according to claim 6, characterized in that, In step S5, based on the resonance peak shift... The corresponding physical displacement level triggers a graded warning: when At that time, a level 2 alert will be output; when At that time, a Level 1 severe warning will be issued; in, and This is the preset frequency offset threshold.

Images