A method for safe operation and maintenance of offshore wind power

Abstract

A kind of offshore wind power safe operation method, comprising the following steps: collecting meteorological data and making prediction, based on predicted wind speed, weather and sea state data, formulate wind turbine work plan;Collecting tower internal gas environment data and ship AIS data, respectively predict the gas release amount and ship arrival time in the future work time;When the ship arrives at the position, judge whether the next 2 hours of wind speed is suitable for stopping and the current sea wave is suitable for boarding and relying;If both are suitable, open the ventilator to ventilate 15 minutes in advance, after the internal gas content is in compliance, trigger voice prompt to open lock.The technical problem to be solved by the present application is to provide a kind of offshore wind power safe operation method, to solve the technical problems of insufficient decision-making ability of environment data driving, lack of multi-source data collaborative control and weak automatic safety guarantee mechanism in existing offshore wind power operation, leading to "high risk, low efficiency and difficult to trace" in operation operation.

Description

Technical Field

[0001] This invention relates to the field of offshore wind power operation and maintenance technology, and in particular to a method for safe operation and maintenance of offshore wind power. Background Technology

[0002] With the rapid growth of global offshore wind power installed capacity, the safe operation and maintenance of offshore wind turbines faces severe challenges. The offshore environment is highly complex (e.g., strong winds, high waves, corrosive gases) and highly uncertain (e.g., rapidly changing weather conditions). Traditional operation and maintenance methods have the following significant drawbacks: 1. Insufficient environmental adaptability Limited application of meteorological data: Existing operation and maintenance plans are usually based on historical meteorological statistics and lack the ability to dynamically respond to real-time meteorological data (such as sudden strong winds and sea fog). This can easily lead to extreme weather during downtime operations, causing equipment damage or personnel safety risks.

[0003] Lack of coordination between sea conditions and ships: The safety of ships berthing at aircraft positions relies on manual observation of parameters such as wave height and ocean currents. There is a lack of trajectory prediction models based on Automatic Identification System (AIS) data, which often leads to efficiency and safety issues such as "excessive waiting time at aircraft positions" or "forced berthing" due to errors in the estimation of ship arrival time.

[0004] 2. Weak safety management inside the tower. Gas monitoring lag: There may be SF6 leaks and accumulation of harmful gases such as hydrogen sulfide inside the tower of offshore wind turbines. However, the existing technology can only achieve "post-event monitoring" of gas concentration and lacks an early warning mechanism based on gas release trends. This can easily lead to maintenance personnel being exposed to danger due to substandard gas environment.

[0005] Inefficient ventilation control: Traditional ventilation fans rely on manual commands to start and stop, and cannot dynamically adjust the ventilation volume according to the gas concentration. This may lead to excessive gas levels due to insufficient ventilation, or energy waste due to excessive ventilation.

[0006] 3. Low level of automation in operation and maintenance processes Data silos across multiple systems: Systems such as meteorological monitoring, gas detection, and ship scheduling operate independently, and the data is not integrated and analyzed, making it difficult to form automated control commands across links (such as automatically adjusting the working plan of the fans based on the weather forecast results, or pre-starting the ventilation process based on the arrival time of ships).

[0007] Safety verification relies on manual labor: Safety verification for critical operations such as shutdown and tower climbing (such as whether the wind speed is suitable for shutdown and whether the gas is compliant) relies on manual inspection and paper records, which is risky for misjudgment and inefficient.

[0008] In summary, existing offshore wind power operation and maintenance methods have significant shortcomings in environmental data-driven decision-making capabilities, multi-source data collaborative control, and automated safety assurance mechanisms, leading to problems of "high risk, low efficiency, and difficulty in traceability" in operation and maintenance. How to achieve "predictive operation and maintenance" and "automated safety control" through technological innovation is a core technical challenge that the industry urgently needs to solve. Summary of the Invention

[0009] The technical problem to be solved by this invention is to provide a method for safe operation and maintenance of offshore wind power, which solves the technical problems of insufficient environmental data-driven decision-making capabilities, lack of multi-source data collaborative control, and weak automated safety assurance mechanisms in the existing offshore wind power operation and maintenance, resulting in "high risk, low efficiency and difficulty in traceability" of operation and maintenance operations.

[0010] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is: a method for safe operation and maintenance of offshore wind power, comprising the following steps: Collect meteorological data and make forecasts; based on the forecasted wind speed, weather and sea state data, formulate wind turbine operation plans. Collect gas environment data inside the tower and ship AIS data to predict the gas release volume and ship arrival time at the engine position during the future working hours, respectively. When the ship arrives at the engine room, determine whether the wind speed in the next 2 hours is suitable for stopping the engine and whether the current sea waves are suitable for landing. If all conditions are met, turn on the ventilation fan 15 minutes in advance. Once the internal gas content is within acceptable limits, trigger a voice prompt to unlock the door.

[0011] Preferably, the meteorological data includes real-time wind speed (accuracy ±0.1m / s), weather conditions (resolution down to the minute level), wave height (measurement error ±5cm), and ocean current speed (measurement error ±0.05m / s). The prediction method uses an LSTM neural network model to make rolling predictions of meteorological parameters for the next 12-24 hours.

[0012] Preferably, the gas environment data inside the tower is collected in real time by an infrared SF6 sensor (detection range 0-2000ppm, accuracy ±5%) and an electrochemical hydrogen sulfide sensor (detection range 0-50ppm, accuracy ±1ppm), and a concentration curve is generated at a frequency of once per second. The trend of gas release in the next 2 hours is predicted by exponential smoothing method.

[0013] Preferably, the ship's AIS data is acquired through a maritime satellite receiving terminal, and the ship's position trajectory is optimized by combining a Kalman filter algorithm. The arrival time at the aircraft position is predicted based on the ship's draft and speed-power curve model, with the error controlled within ±5 minutes.

[0014] Preferably, the criteria for determining whether the wind speed is suitable for shutdown are: predicted wind speed ≤8m / s (for 3MW and above models) or ≤10m / s (for 2MW and below models), and wind speed fluctuation rate ≤±20% in the next 4 hours.

[0015] Preferably, the criteria for determining whether the waves are suitable for boarding are: effective wave height ≤ 1.2m (applicable to monohull maintenance vessels) or ≤ 1.5m (applicable to catamaran maintenance vessels), ocean current speed ≤ 0.8kn, and the ship's roll angle ≤ 8° and pitch angle ≤ 5°.

[0016] Preferably, the ventilation volume of the ventilator is dynamically adjusted according to the tower volume and gas concentration. The initial ventilation power is ≥15kW, the air exchange rate is ≥15 times / hour, and the fan speed is adjusted in real time by a PID controller during ventilation. When the SF6 concentration is ≤800ppm and the hydrogen sulfide concentration is ≤5ppm (the three consecutive detection values ​​meet the standard), the fan automatically switches to a low-speed maintenance mode (power ≤5kW) and triggers a gas compliance signal.

[0017] Preferably, the voice prompt unlocking must meet three confirmation conditions: ① gas compliance signal (cross-verification by dual sensors); ② the error between the ship's GPS position and the aircraft position coordinates is ≤50m; ③ the maintenance personnel send a tower climbing confirmation command via a handheld terminal, and the voice content includes the operation time and safety precautions.

[0018] Preferably, the wind turbine work plan is managed using a Gantt chart visualization, which includes a flexible time window of ±2 hours. When the predicted meteorological parameters approach the threshold (such as wind speed reaching 80% of the shutdown threshold), a plan adjustment warning is automatically triggered, and the work sequence is re-optimized through a multi-objective genetic algorithm.

[0019] Preferably, the collected data adopts a collaborative architecture of edge computing and cloud computing: At the edge, real-time meteorological data, gas concentrations, and ship trajectories are preprocessed (filtering and noise reduction, outlier marking) and transmitted to the cloud via a 5G network; A digital twin model is established in the cloud to simulate the gas diffusion path inside the tower and the risk areas of ship navigation in real time, and a dynamic risk assessment report is generated based on a Bayesian network. Data storage uses a distributed file system (HDFS), which supports second-level data backtracking and multi-dimensional comparative analysis. Abnormal data automatically triggers the work order system, generates hidden danger rectification tasks, and associates them with the operation and maintenance personnel's accounts.

[0020] This invention provides a method for safe operation and maintenance of offshore wind power. By collecting meteorological, tower internal gas environment, and ship AIS data and using algorithms such as LSTM neural networks and Kalman filters for prediction, it dynamically formulates wind turbine operation plans. It sets differentiated judgment criteria such as wind speed (≤8m / s or ≤10m / s) and wave height (≤1.2m or ≤1.5m) based on turbine type (3MW and above / 2MW and below) and ship type (monoplanet / catamaran), enabling intelligent judgment of environmental conditions. It automatically starts the variable frequency centrifugal fan 15 minutes in advance and dynamically adjusts the ventilation volume. Voice prompts for unlocking are triggered by triple confirmation conditions. A Gantt chart is used for visual management of the plan, with flexible time windows and automatic adjustment of early warnings. An edge computing and cloud computing collaborative architecture is used to achieve data preprocessing, digital twin simulation, dynamic risk assessment, and distributed storage traceability. It has advantages such as strong environmental adaptability, precise tower gas risk control, high degree of automation in operation and maintenance processes, and scientific and efficient data-driven decision-making. It can improve the safety, efficiency, and intelligence level of operation and maintenance, reduce operation and maintenance costs, and achieve closed-loop management of potential hazards. Attached Figure Description

[0021] The present invention will be further described below with reference to the accompanying drawings and embodiments: Figure 1 This is a flowchart of the method of the present invention. Detailed Implementation

[0022] like Figure 1 As shown, a method for safe operation and maintenance of offshore wind power includes the following steps: Collect meteorological data and make forecasts; based on the forecasted wind speed, weather and sea state data, formulate wind turbine operation plans. Collect gas environment data inside the tower and ship AIS data to predict the gas release volume and ship arrival time at the engine position during the future working hours, respectively. When the ship arrives at the engine room, determine whether the wind speed in the next 2 hours is suitable for stopping the engine and whether the current sea waves are suitable for landing. If all conditions are met, turn on the ventilation fan 15 minutes in advance. Once the internal gas content is within acceptable limits, trigger a voice prompt to unlock the door.

[0023] Full-process closed-loop control: Constructing a complete chain of "data acquisition → predictive analysis → condition judgment → execution control", covering the core links of offshore wind power operation and maintenance (meteorological response, gas management, and ship coordination), and realizing automated linkage from environmental perception to safe operation.

[0024] Multi-source data fusion: Integrating three key data types—meteorological, gas, and ship data—breaks through the limitations of traditional "single data-driven" operation and maintenance, and improves the scientific nature of decision-making.

[0025] Preferably, the meteorological data includes real-time wind speed (accuracy ±0.1 m / s), weather conditions (resolution down to the minute level), wave height (measurement error ±5 cm), and ocean current speed (measurement error ±0.05 m / s). The prediction method employs an LSTM neural network model to perform rolling forecasts of meteorological parameters for the next 12-24 hours. Using an LSTM neural network for 12-24 hour rolling forecasts captures the dynamic trends of meteorological parameters and is more adaptable to the high uncertainty of marine weather compared to traditional statistical methods, reducing the risk of misjudging extreme weather conditions.

[0026] Preferably, the gas environment data inside the tower is collected in real time using an infrared SF6 sensor (detection range 0-2000ppm, accuracy ±5%) and an electrochemical hydrogen sulfide sensor (detection range 0-50ppm, accuracy ±1ppm). A concentration curve is generated at a frequency of once per second, and the gas release trend for the next two hours is predicted using an exponential smoothing method. Predicting the gas release for the next two hours using exponential smoothing transforms "passive detection" into "active early warning," allowing for advance planning of ventilation pretreatment and ensuring the safety of maintenance personnel.

[0027] Preferably, the ship's AIS data is acquired through a maritime satellite receiving terminal. The ship's position trajectory is optimized using a Kalman filter algorithm, and the arrival time at the aircraft stand is predicted based on the ship's draft and speed-power curve model, with the error controlled within ±5 minutes. By using the Kalman filter algorithm to eliminate noise in the ship's AIS data and combining it with the ship's dynamics model (draft, speed-power curve), the arrival time prediction error is controlled within ±5 minutes, reducing "waiting at the aircraft stand" or "delayed boarding" and improving the efficiency of the operation and maintenance process.

[0028] Preferably, the criteria for determining whether the wind speed is suitable for shutdown are: predicted wind speed ≤8m / s (for 3MW and above models) or ≤10m / s (for 2MW and below models), and wind speed fluctuation rate ≤±20% in the next 4 hours.

[0029] Preferably, the criteria for determining whether the waves are suitable for boarding are: effective wave height ≤ 1.2m (applicable to monohull maintenance vessels) or ≤ 1.5m (applicable to catamaran maintenance vessels), ocean current speed ≤ 0.8kn, and the ship's roll angle ≤ 8° and pitch angle ≤ 5°.

[0030] Preferably, the ventilation volume of the ventilator is dynamically adjusted according to the tower volume and gas concentration, with an initial ventilation power ≥15kW and an air exchange rate ≥15 times / hour. During ventilation, the fan speed is adjusted in real time by a PID controller. When the SF6 concentration is detected to be ≤800ppm and the hydrogen sulfide concentration is ≤5ppm (three consecutive detection values ​​meet the standards), the system automatically switches to a low-speed maintenance mode (power ≤5kW) and triggers a gas compliance signal. This avoids the problems of "energy waste" or "insufficient ventilation" associated with traditional ventilation systems.

[0031] Preferably, the voice prompt unlocking requires three confirmation conditions: ① gas compliance signal (dual sensor cross-verification); ② ship GPS position and aircraft position coordinate error ≤ 50m; ③ maintenance personnel send a tower access confirmation command via handheld terminal, with the voice content including operation time and safety precautions. This triple safety verification—gas compliance (dual sensor cross-verification), ship position error ≤ 50m, and manual terminal confirmation—forms a "equipment-environment-personnel" triple insurance, preventing accidental unlocking due to a single condition failure and improving operational safety.

[0032] Preferably, the wind turbine work plan is managed using a Gantt chart visualization, including a ±2-hour flexible time window. When predicted meteorological parameters approach a threshold (e.g., wind speed reaches 80% of the shutdown threshold), an automatic plan adjustment warning is triggered, and the work sequence is re-optimized through a multi-objective genetic algorithm. Visualization and Flexible Management: The Gantt chart intuitively displays the plan time window (±2-hour flexibility), facilitating resource coordination by the operations and maintenance team; automatic warnings are issued when thresholds are approached, and the work sequence is optimized through a multi-objective genetic algorithm, dynamically responding to weather changes and reducing plan delays.

[0033] Intelligent scheduling: It replaces traditional manual scheduling, improves the efficiency of multi-task collaboration, and is especially suitable for multi-station operation and maintenance scenarios under complex weather conditions.

[0034] Preferably, the collected data adopts a collaborative architecture of edge computing and cloud computing: At the edge, real-time meteorological data, gas concentrations, and ship trajectories are preprocessed (filtering and noise reduction, outlier marking) and transmitted to the cloud via a 5G network; A digital twin model is established in the cloud to simulate the gas diffusion path inside the tower and the risk areas of ship navigation in real time, and a dynamic risk assessment report is generated based on a Bayesian network. Data storage uses a distributed file system (HDFS), which supports second-level data backtracking and multi-dimensional comparative analysis. Abnormal data automatically triggers the work order system, generates hidden danger rectification tasks, and associates them with the operation and maintenance personnel's accounts.

[0035] Real-time data preprocessing (filtering and noise reduction) at the edge reduces network transmission pressure, while 5G transmission ensures real-time performance; cloud-based digital twins simulate gas diffusion and ship risks, and Bayesian networks generate dynamic risk reports, achieving intelligent operation across the entire chain of "data acquisition - risk prediction - decision support".

[0036] Distributed storage (HDFS) supports second-level backtracking and multi-dimensional analysis. Abnormal data automatically triggers the work order system, forming a closed loop of "monitoring-analysis-rectification" to improve the standardization of operation and maintenance and the efficiency of accident tracing.

[0037] During operation, meteorological data (including real-time wind speed, weather conditions, wave height, ocean current speed, etc.), internal gas environment data of the tower (SF6 and hydrogen sulfide concentrations are collected once per second via infrared SF6 sensors and electrochemical hydrogen sulfide sensors), and ship AIS data (obtained via a maritime satellite receiving terminal) are collected first. Then, LSTM neural network models, exponential smoothing, and Kalman filtering algorithms combined with ship dynamics models are used to predict future meteorological parameters, gas release trends, and ship arrival time at the turbine position (error ±5 minutes). Based on the prediction results, a wind turbine operating plan with a ±2-hour flexible window is formulated using a Gantt chart. When the ship arrives at the turbine position, the wind speed (≤8m / s or ≤10m / s with fluctuation rate ≤±20%) and current wave conditions (significant wave height ≤1.2m or ≤1.5m, ocean current speed ≤0.8kn, roll angle ≤8°, pitch angle ≤5°) for the next 2 hours are assessed based on the turbine type (3MW and above / 2MW and below) and ship type (monocoque / catamaran). If all conditions are suitable, the wind turbine is deployed 15 minutes in advance. The variable frequency centrifugal fan (initial power ≥15kW, air exchange rate ≥15 times / hour, speed dynamically adjusted by PID controller) is activated within minutes. After three consecutive tests showing SF6 concentration ≤800ppm and hydrogen sulfide concentration ≤5ppm, a voice prompt to unlock is triggered after meeting three conditions: gas compliance signal, ship GPS position and aircraft position coordinate error ≤50m, and confirmation by maintenance personnel's handheld terminal. During operation, data is preprocessed at the edge and transmitted to the cloud via 5G. A digital twin model is used to simulate gas diffusion and ship risk areas. A risk report is generated based on a Bayesian network. Data is stored in a distributed file system (HDFS) and supports second-level backtracking, multi-dimensional analysis, and abnormal triggering of work orders to rectify potential hazards.

[0038] This invention collects meteorological, tower internal gas environment, and ship AIS data, and uses algorithms such as LSTM neural networks and Kalman filters to predict wind speed, gas release, and ship arrival time. It sets differentiated judgment criteria based on aircraft and ship types, including wind speed (≤8m / s or ≤10m / s) and wave height (≤1.2m or ≤1.5m). When environmental conditions are suitable, it automatically starts the variable frequency centrifugal fan 15 minutes in advance and dynamically adjusts the ventilation volume. After meeting three conditions—gas compliance, ship positioning error ≤50m, and manual confirmation—it triggers a voice prompt to unlock the door. It uses a Gantt chart for visual management of work plans with flexible windows, and utilizes a collaborative architecture of edge computing and cloud computing to achieve data preprocessing, digital twin simulation, Bayesian network risk assessment, and distributed storage traceability. This improves the safety, efficiency, and intelligence of operation and maintenance, achieving closed-loop management of potential hazards.

[0039] The above embodiments are merely preferred technical solutions of the present invention and should not be considered as limitations on the present invention. The scope of protection of the present invention should be limited to the technical solutions described in the claims, including equivalent substitutions of the technical features described in the claims. That is, equivalent substitutions and improvements within this scope are also within the scope of protection of the present invention.

Claims

1. A method for safe operation and maintenance of offshore wind power, characterized in that, Includes the following steps: Collect meteorological data and make forecasts; based on the forecasted wind speed, weather and sea state data, formulate wind turbine operation plans. Collect gas environment data inside the tower and ship AIS data to predict the gas release volume and ship arrival time at the engine position during the future working hours, respectively. When the ship arrives at the engine room, determine whether the wind speed in the next 2 hours is suitable for stopping the engine and whether the current sea waves are suitable for landing. If all conditions are met, turn on the ventilation fan 15 minutes in advance. Once the internal gas content is within acceptable limits, trigger a voice prompt to unlock the door. The meteorological data includes real-time wind speed, weather conditions, wave height, and ocean current speed. The prediction method uses an LSTM neural network model to make rolling predictions of meteorological parameters for the next 12-24 hours. The gas environment data inside the tower is collected in real time by an infrared SF6 sensor and an electrochemical hydrogen sulfide sensor. Concentration curves are generated at a frequency of once per second, and the gas release trend in the next 2 hours is predicted by exponential smoothing. The criteria for determining whether the wind speed is suitable for shutdown are: predicted wind speed ≤ 8 m / s, and wind speed fluctuation rate ≤ ±20% in the next 4 hours; The ventilation volume of the ventilator is dynamically adjusted according to the tower volume and gas concentration. The initial ventilation power is ≥15kW and the air exchange rate is ≥15 times / hour. During the ventilation process, the fan speed is adjusted in real time by a PID controller. When the SF6 concentration is ≤800ppm and the hydrogen sulfide concentration is ≤5ppm, the fan automatically switches to low-speed maintenance mode and triggers a gas compliance signal. The voice prompt for unlocking requires three confirmation conditions: ① gas compliance signal; ② the error between the ship's GPS position and the aircraft position coordinates is ≤50m; ③ the maintenance personnel send a confirmation command for climbing the tower via a handheld terminal, and the voice content includes the operation time and safety precautions. The wind turbine work plan is managed visually using a Gantt chart. When the predicted meteorological parameters approach the threshold, an early warning for plan adjustment is automatically triggered, and the work sequence is re-optimized through a multi-objective genetic algorithm.

2. The method for safe operation and maintenance of offshore wind power according to claim 1, characterized in that, The ship's AIS data is acquired through a maritime satellite receiving terminal. The ship's position trajectory is optimized by combining the Kalman filter algorithm, and the arrival time at the aircraft position is predicted based on the ship's draft and speed-power curve model.

3. The method for safe operation and maintenance of offshore wind power according to claim 1, characterized in that, The criteria for determining whether a ship is suitable for boarding are: a significant wave height ≤ 1.2m, a current speed ≤ 0.8kn, and a roll angle ≤ 8° and a pitch angle ≤ 5°.

4. The method for safe operation and maintenance of offshore wind power according to claim 1, characterized in that, The collected data adopts a collaborative architecture of edge computing and cloud computing: The edge device preprocesses real-time collected meteorological data, gas concentrations, and ship trajectories, and transmits them to the cloud via a 5G network; A digital twin model is established in the cloud to simulate the gas diffusion path inside the tower and the risk areas of ship navigation in real time, and a dynamic risk assessment report is generated based on a Bayesian network. Data storage uses a distributed file system, supporting second-level data backtracking and multi-dimensional comparative analysis. Abnormal data automatically triggers the work order system, generating hidden danger rectification tasks and associating them with the operation and maintenance personnel's accounts.

Images