Steady Power Output in Wind Turbines

Short- to medium-term wind power prediction method and system for wind farms that improves accuracy by using real-time blade parameters and historical data from multiple wind turbines. Each wind turbine has an edge controller with shared memory. The controller collects blade parameters, predicts local wind power, and stores historical data. Multiple algorithms are called to predict wind power using real-time data and historical data from nearby turbines. The predictions are weighted and summed for the final result.

Source

Dual-mode control method and system for timely monitoring of wind turbines based on wind speed to improve wind power generation efficiency. The method involves using machine learning to predict wind speed trends, analyze factors affecting wind speed changes, and quantify wind speed variability. This allows timely adjustment of wind turbine operation to optimize power generation in response to changing wind conditions.

Source

Intelligent real-time monitoring system and method for wind turbines that optimizes power generation and reduces failures by actively adjusting the blade orientation to track the wind direction. The system uses sensors to detect wind direction and blade position, calculates the blade angle relative to wind, and uses a stepper motor to rotate the blade assembly to align with the wind. This allows the turbine to better capture wind energy and avoid stalling. The system also monitors blade forces and adjusts orientation based on thresholds to prevent overloading.

Source

Optimizing wind farm active power regulation during grid frequency fluctuations using model predictive control to accurately allocate adjustment amounts to each wind turbine. This involves quantifying the active support capability of wind turbines using simulation when they only use rotor kinetic energy for short-term regulation. The turbine's maximum up and down power adjustments are calculated under different wind speeds. Then during frequency deviations, the total farm adjustment is computed and allocated per turbine via MPC to avoid exceeding their support limits and prevent rotor speed issues.

Source

Intelligent monitoring and prediction system for wind power towers to improve efficiency and prevent issues. The system uses sensors like compass, anemometer, inclinometer, temperature/humidity on each tower, along with a central control board, 4G module, cloud server, Beidou module, and wind turbine controller. It predicts overall wind direction and strength for the tower group using data from multiple towers, allowing advance control to optimize energy capture. This improves efficiency compared to single tower prediction. It also detects issues like tower tilt or vibration, allowing early warning and response.

Source

Robust adaptive fault-tolerant control method for wind turbines using machine learning to improve fault detection and adaptation for wind turbines. The method involves training a machine learning model using historical wind turbine data to learn normal operating conditions and failures. During real-time operation, the model is used to continuously monitor the wind turbine and detect deviations from normal behavior. If a fault is detected, the model provides recommendations for adaptive control strategies to mitigate the fault and maintain stability. The adaptive control is based on the specific fault type and severity, learned from the training data. This allows the wind turbine to respond dynamically to unforeseen failures and environmental changes, improving fault tolerance and system stability compared to traditional control methods.

Source

Power generation control method for wind turbines using a SCADA system to optimize performance and reduce failures. The method involves monitoring wind turbine equipment, generating expected power based on wind speed, comparing actual power, and adjusting parameters if needed. The comparison uses evaluation coefficients to determine when to correct. If the actual power deviates significantly from expected, a correction is generated. If the deviation is smaller, a trend analysis is done to decide if a correction is needed. This allows targeted adjustments to wind turbine parameters based on actual operating conditions.

Source

Method for evaluating the transient active power regulation capability of wind turbines using rotor kinetic energy control. It provides a rapid and accurate way to assess how much active power a wind farm can provide during grid disturbances. The method involves optimizing wind turbine control instructions to maximize active power support while considering rotor speed, torque, and inverter capacity constraints. An optimization algorithm is used to find the optimal adjustment amounts and response times. This allows rapid calculation of the wind farm's transient active power support capacity.

Source

Optimizing the droop coefficient of wind turbines using model predictive control to improve frequency regulation and stability of wind power systems. The method involves collecting system data, building a frequency response model, setting an initial droop value, iteratively optimizing the droop through model predictive control, and simulating time domain operation to find the optimal sequence of droop coefficients for wind turbines during frequency disturbances.

Source

Optimizing wind turbine operation to balance energy production, grid stability, and maintenance scheduling. The method involves forecasting operating parameters like turbine health and wind conditions, then determining a control scheme for the turbine that balances factors like energy production, failure risk, and grid demand. It uses evaluation matrices and optimization techniques to find the best timing and power limits for shutdowns and operation. This balances factors like energy production, grid stability, and maintenance scheduling.

Source

Method and system for switching operating states of wind turbines that accurately predicts and controls turbine operation for improved efficiency, fault handling, and wind load mitigation. The method involves collecting turbine and wind data, determining status modes like startup, generation, shutdown, etc., predicting optimal adjustments based on models, and regulating blade angle and speed using a PID controller. This allows targeted response to wind conditions and faults rather than blanket shutdowns.

Source

Online monitoring of wind turbine power using machine learning to improve reliability and timeliness compared to existing methods. The method involves training a wind speed-power mapping model using historical data, calculating power discreteness in wind speed intervals, and using that to assess wind turbine power normality in real time. This allows quick detection of abnormal power and reduces losses compared to waiting for periodic offline monitoring.

Source

Controlling the rate of change of power in wind farms using inertia response of wind turbines to smooth power output. The method involves monitoring the rate of change of active power in the wind farm, and independently controlling each wind turbine's power change rate based on its inertia response. This reduces turbine-level power rate changes. If the wind farm-wide power rate change exceeds a threshold, coordinated control is triggered to further reduce the overall power rate change. By leveraging turbine inertia, it smoothens wind farm power output compared to just changing power quickly.

Source

Wind turbine control method that mitigates oscillations and instability caused by fluctuating wind speeds. It considers short-term fluctuations in wind speed and generator speed to determine if the turbine dynamics will be affected. By calculating an index based on these fluctuations, the method can instruct the main turbine control system to change pitch or shut down to prevent oscillations. This prevents dynamic issues caused by imperfect logic when wind speeds fluctuate greatly.

Source

Intelligent control of wind turbines that can accurately predict and manage the load conditions of the turbine blades under different wind conditions. The method involves training a neural network model using weighted parameter distributions to accurately predict blade load for different wind conditions. This allows safe operation of the turbines by intelligently controlling them based on predicted loads. The weights are determined by analyzing correlation between wind parameters and load for different blade positions.

Source

Optimizing wind turbine performance during noise reduced operation to maximize energy production without accelerating component wear. The method involves dynamically adjusting turbine power output based on grid conditions instead of fixed noise reduction modes. When prioritizing component life, turbine limits power near sync speed to prevent grid issues. When prioritizing energy, it operates at max rating tracking component degradation. This allows higher output without consuming life faster.

Source

A wind farm power control system that allows wind turbines to respond to grid dispatch commands for better grid stability. The system involves monitoring the actual power output of each turbine and using a central controller to generate power change signals based on grid commands. These signals are sent to the turbine control modules to adjust turbine parameters and optimize output. This allows the turbines to dynamically respond to grid requirements and improve wind farm power regulation.

Source

Method for monitoring and controlling wind turbine blades without modifying them. It involves scanning the blade shapes using lasers and processing the real-time data to identify the blade angles accurately. This allows determining the blade state without needing to physically measure the angles. The blades can be scanned periodically using lasers added to the turbine. The scanned data is processed using algorithms to accurately determine the blade angles. This is used to optimize blade positioning for maximum power generation and detect blade issues like deformation or imbalance.

Source

Model predictive control (MPC) of wind turbines that is robust to gusts and parametric variations in operational parameters. The MPC algorithm uses a wind turbine model to predict future states and optimize control actions over a horizon. To match the sampling rate of measured sensor inputs to the predicted values, the MPC calculation rate is lower than the sensor sampling rate. This ensures accurate timing between measured and predicted values. This allows using higher sensor sampling rates than the MPC calculation rate to improve accuracy. The lower MPC calculation rate allows using a more complex wind turbine model with parameters like aerodynamic thrust and power, rather than just rotor speed.

Source

Optimizing wind turbine performance and control using physics-based modeling and machine learning. The method involves leveraging detailed wind turbine models to estimate aerodynamic states and operating conditions in real time. It uses multiple physics-based aeroelastic estimators and machine learning algorithms to accurately determine wind resource and fault conditions based on operational data. The estimator configurations and control strategies are adaptively configured based on the estimated states to provide optimal performance and fault management. This enables full utilization of aeroelastic predictive control for core turbine functions.

Source