Validate DFIG Wind Turbine Performance Under Grid Voltage Sag

Overview of Technical Issues:

During grid voltage sag events, the rotor current control unit insufficiently constrains transient overcurrents because voltage drops induce electromagnetic flux imbalances that generate currents exceeding rated values within milliseconds; simultaneously, the electromagnetic torque generation unit experiences harmful oscillations from unbalanced flux linkage, creating mechanical stress and potential converter thermal damage; these functional defects cause the DFIG to fail low voltage ride-through requirements and disconnect from the grid during faults; the goal is to validate and achieve stable operation with the turbine remaining connected during voltage sags down to 0.15 per unit for durations specified by grid codes.

Solution directions generated for this problem

Problem Direction 1 :

ImproveRotor current control response speed
VS
ConstraintConverter thermal loading

Inspiration 1 : Cross-domain reference

Application Principle: #19 Periodic action
Cross-domain applicability Assess applicability
Motor control device and brake fluid pressure device for a vehicle
Innovative Solution Refine solution

Burst-mode adaptive switching for transient current control

Cyclic burst-mode switching during voltage sag
How to solve :
  • Implement three-phase burst-mode switching: operate converter at maximum frequency 8kHz for initial 3ms when voltage sag detected below 0.5pu to achieve sub-millisecond current response, then reduce to 2kHz once rotor current constrained below 1.8pu
  • Deploy voltage sag depth detector with 50μs sampling to trigger burst cycles — deeper sags (0.15-0.3pu) maintain 8kHz for 5ms, moderate sags (0.3-0.6pu) for 3ms, shallow sags (0.6-0.9pu) for 1ms before frequency reduction
  • Install real-time thermal monitor measuring IGBT junction temperature via integrated NTC sensors with ±2°C accuracy, automatically extend low-frequency period if temperature exceeds 105°C during fault
Expected Effect : Switching loss reduced 40-55% vs continuous high-frequency operation; sub-millisecond response maintained; LVRT success rate ≥98% for 0.15pu sag
Risk Control :
  • burst timing calibration error causing inadequate current limiting
  • thermal sensor lag delaying protection
  • transition transient between frequency modes inducing voltage spikes

Problem Direction 2 :

ImproveRotor current control response speed
VS
ConstraintControl system complexity

Inspiration 1 : Cross-domain reference

Application Principle: #26 Copying
Cross-domain applicability Assess applicability
System and method for generating recommendations during image capture of a product
Innovative Solution Refine solution

Pre-computed lookup table for sub-millisecond rotor current control during voltage sag

Replace real-time computation with pre-calculated control actions stored in lookup tables
How to solve :
  • Pre-calculate optimal rotor current control actions for a 3D lookup table indexed by voltage sag depth (0.15–1.0 pu in 0.05 pu steps), phase angle (0–360° in 15° steps), and rotor speed (0.7–1.3 pu in 0.1 pu steps), storing d-axis and q-axis current references
  • total table size 2 MB, stored in fast SRAM for sub-100 microsecond retrieval
  • Implement trilinear interpolation between nearest table entries using fixed-point arithmetic (16-bit precision) to achieve smooth control transitions with ±2% accuracy compared to real-time calculation, executed in 50 microseconds on standard industrial DSP (TMS320F28379D)
  • Deploy voltage sag detector using three-sample moving window (sampling at 10 kHz) to trigger table lookup mode within 300 microseconds of sag initiation, bypassing complex field-oriented control equations and flux observers during transient period
Expected Effect : Control response <500 μs; computational load −85%; rotor overcurrent limited to 1.8 pu
Risk Control :
  • lookup table coverage gaps in edge cases
  • interpolation accuracy degradation at boundaries
  • memory integrity under electromagnetic interference

Problem Direction 3 :

ImproveElectromagnetic flux imbalance damping capability
VS
ConstraintControl system complexity

Inspiration 1 : Cross-domain reference

Application Principle: #35 Parameter changes
Cross-domain applicability Assess applicability
DC bus voltage control
Innovative Solution Refine solution

Adaptive rotor circuit impedance modulation for passive flux damping

Modify rotor circuit physical properties for passive damping
How to solve :
  • Install variable inductance module (0.8–2.5 mH range) in series with rotor windings, switched via thyristor bypass circuit triggered by grid voltage threshold detector set at 0.85 per unit
  • During normal operation, thyristor conducts to bypass the inductance (total added impedance <0.3 mΩ)
  • upon voltage sag detection within 0.5 ms, thyristor blocks and inserts the inductance, creating natural current-limiting at 1.8 per unit and passive eddy-current damping that suppresses flux oscillations by ≥60% without control algorithm modification
  • Quality control: inductance tolerance ±5%, thyristor turn-off time ≤50 μs verified by oscilloscope, temperature rise under 2.0 pu rotor current ≤40°C measured by embedded PT100 sensors
  • Implementation: mount air-core inductor (litz wire, 200 strands × 0.2 mm) in rotor-side cabinet, parallel thyristor module rated 1200V/600A with RC snubber (47 Ω, 0.1 μF), analog comparator (LM339) monitors stator voltage via potential transformer for threshold detection
Expected Effect : Flux oscillation amplitude reduced 60–75%; zero added computational load; hardware cost <$800 per unit; LVRT compliance to 0.15 pu for 625 ms
Risk Control :
  • thyristor commutation failure during deep sag
  • inductance thermal drift affecting damping coefficient
  • EMI from fast thyristor switching

Problem Direction 4 :

ImproveGrid connection stability during voltage sag
VS
ConstraintConverter thermal loading

Inspiration 1 : Cross-domain reference

Application Principle: #11 Beforehand cushioning
Cross-domain applicability Assess applicability
Method for depositing an organoaminosilane precursor and a thin film containing the precursor
Innovative Solution Refine solution

Pre-fault thermal margin reservation with predictive grid disturbance monitoring for LVRT

Pre-fault thermal reserve via grid monitoring
How to solve :
  • Deploy grid disturbance predictor monitoring voltage harmonics (THD threshold 3%) and frequency deviation (±0.2 Hz) 500–2000 ms before fault onset
  • upon detection, activate pre-cooling protocol: increase converter fan speed from 40% to 100%, reduce rotor current reference by 12–18% for 1–3 seconds to lower IGBT junction temperature by 8–15°C below rated 125°C
  • During voltage sag (0.15–0.9 pu), utilize the thermal headroom created to sustain 1.8–2.2 pu rotor current for grid code duration (150–625 ms) without exceeding 135°C thermal limit
  • Implement adaptive thermal budget controller: real-time junction temperature estimation via NTC sensors (±2°C accuracy, 50 ms sampling), dynamically allocate current capacity based on remaining thermal margin, ensuring LVRT completion with 10°C safety buffer
Expected Effect : LVRT success rate 98%, thermal margin +15°C, no added hardware
Risk Control :
  • false disturbance prediction causing unnecessary derating
  • thermal model accuracy degradation over component aging
  • grid code duration exceeding thermal budget reserve

Problem Direction 5 :

ImproveGrid connection stability during voltage sag
VS
ConstraintControl system complexity

Inspiration 1 : Cross-domain reference

Application Principle: #1 Segmentation
Cross-domain applicability Assess applicability
Billing usage in a virtual computing infrastructure
Innovative Solution Refine solution

Modular plug-in LVRT controller with dormant-active state switching

Separate LVRT into plug-in module
How to solve :
  • Implement LVRT as a standalone plug-in module with dedicated 32-bit microcontroller (ARM Cortex-M4, 168MHz) that remains dormant during normal operation and activates only when grid voltage drops below 0.9 per unit, isolating fault-handling complexity from main control system
  • The plug-in module executes only three hardcoded functions: rotor current limiting to 2.0 pu via lookup table (pre-calculated for voltage sag depths 0.15-1.0 pu in 0.05 pu steps), DC-link voltage damping via single high-pass filter (cutoff 10Hz, gain -0.3), and synchronization maintenance using phase-locked loop with fixed bandwidth 15Hz
  • Module interfaces with main controller through four discrete signals (fault detect, takeover request, current limit value, damping signal) and returns control within 50ms after voltage recovers above 0.85 pu for 20ms continuous duration, verified by dual-threshold comparator with 2ms debounce time
Expected Effect : LVRT success rate ≥99.5% for 0.15pu sag; main control code +8% only; commissioning time -40%
Risk Control :
  • plug-in module communication latency exceeding 5ms
  • lookup table boundary interpolation error
  • state handover transient causing current spike
Patsnap Eureka Solution