The online monitoring method of the transformer comprehensive testing system achieves comprehensive perception of equipment status and fault warning through real-time collection and intelligent analysis of multi-dimensional parameters. The core monitoring methods and technical points are as follows:
1、 Core monitoring parameters and sensor technology
Electrical parameter monitoring
Voltage/current monitoring: Real time collection of input and output parameters through current sensors and voltage sensors, combined with power analyzers to calculate active/reactive power and determine transformer load status.
Frequency monitoring: detecting fluctuations in the power grid frequency to prevent equipment loss caused by abnormal frequency.
Ratio test: Verify the consistency between the actual ratio of the transformer and the design value to ensure normal voltage conversion capability.
Oil system monitoring
Dissolved Gas Analysis (DGA) in Oil: Using gas chromatography or infrared spectroscopy techniques, real-time analysis of gas concentrations such as H ₂, CH ₄, C ₂ H ₂, etc., to identify fault types such as overheating and discharge. For example, excessive acetylene (C ₂ H ₂) may indicate arc discharge.
Oil temperature monitoring: Real time tracking of oil temperature through fluorescent fiber optic temperature sensors, combined with top-level oil temperature data to establish a dynamic load capacity model to prevent thermal runaway.
Oil quality analysis: detecting parameters such as moisture, acid value, and metal particles in the oil to evaluate the degree of oil degradation. For example, a moisture content exceeding 100ppm will accelerate insulation aging.
Mechanical condition monitoring
Vibration analysis: Use MEMS vibration sensors to collect mechanical vibration signals (10-1000Hz) of iron cores and windings, and combine voiceprint features to identify defects such as looseness and deformation.
Noise monitoring: Collecting 20-200kHz ultrasonic waves through a voiceprint online monitoring system, combined with AI algorithms to distinguish discharge types (such as corona discharge and suspended discharge), with strong anti-interference ability.
Insulation performance monitoring
Partial discharge monitoring:
High frequency current method (HFCT): Install a high-frequency sensor on the grounding wire to capture discharge signals ranging from 300kHz to 30MHz and locate internal defects.
Ultrasonic method: using a magnetic adsorption ultrasonic sensor to adsorb on the outer wall of the oil tank and receive the ultrasonic waves generated by discharge (main frequency 30-180kHz), suitable for oil immersed transformers.
Ultra high frequency (UHF) method: detects electromagnetic waves ranging from 300MHz to 3GHz, with high sensitivity, suitable for GIS transformers.
Monitoring of dielectric loss factor (tan δ): Online monitoring of the tan δ value and capacitance changes of the casing to detect moisture or delamination defects.
Core grounding current monitoring: Real time monitoring of grounding current amplitude and harmonic characteristics through high-precision through type current sensors to diagnose insulation degradation or magnetic saturation hazards in the core.
2、 Intelligent analysis and warning technology
Multi parameter fusion analysis
Integrate 20+parameters such as oil chromatography, partial discharge, temperature, etc., upload them to the main station through the IEC61850 protocol, and establish a prediction model using random forest and LSTM algorithms. For example, a 500kV transformer can provide a 72 hour advance warning for inter turn short circuit faults through the fusion analysis of oil chromatography and partial discharge data.
Edge computing and Localization Decision
The monitoring terminal is equipped with an edge analysis module to achieve localized fault classification and reduce data transmission latency. For example, the FG-BYQ comprehensive online monitoring system has been rigorously tested in the laboratory, and the operational site runs stably and reliably.
Digital Twin and Simulation Technology
Combining 3D modeling to achieve state simulation, simulating the operating state of transformers under different working conditions, and optimizing maintenance strategies.
3、 Typical application scenarios and advantages
power system
Real time monitoring of the operating status of ultra-high voltage transformers to improve power supply reliability. For example, a 500kV substation of State Grid achieved an accuracy rate of over 95% in fault warning through a comprehensive online monitoring system.
industrial and mining enterprises
Monitor industrial power transformers to improve equipment safety and operational efficiency. For example, a certain steel company detected loose winding defects in advance through vibration and noise monitoring, avoiding unplanned downtime losses.
new energy sector
Develop specialized gas monitoring solutions for new energy station transformers, such as hydrogen fuel cell monitoring technology, to meet complex environmental requirements.
4、 Technological development trends
Improved anti-interference ability of sensors
Develop new sensing technologies (such as fluorescent fiber temperature measurement and MEMS vibration sensors) to improve signal recognition accuracy.
Development of Multi physics Field Coupling Analysis Model
Establish a more accurate fault prediction model by combining multidimensional parameters such as electrical, mechanical, and thermal.
Economic plan for monitoring and retrofitting old equipment
Develop low-cost and easy to install monitoring modules for old transformers to promote the popularization of condition based maintenance.