Offshore wind power intelligent auxiliary data processing method

Abstract

The invention relates to the technical field of data processing, in particular to an intelligent auxiliary data processing method for offshore wind power. The method comprises the following steps: performing data acquisition by using the fan auxiliary detection system and the offshore booster station monitoring system respectively; performing data transmission on the acquired data based on the station side and the station-to-center side; wherein in the transmission process, an AES algorithm is adopted for encryption; carrying out data storage on the encrypted data; and carrying out data processing on the stored data through the gateway system. By implementing the intelligent auxiliary data processing method, the cost can be significantly reduced. For example, a high-precision sensor and a data preprocessing technology can capture and process key operation data in time, so that the probability of fault occurrence and the number of times of maintenance are reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and in particular to an offshore wind power intelligent auxiliary data processing method. Background Art

[0002] The operation and maintenance costs of offshore wind power are high, accounting for about 25% of the total project cost. At the same time, the failure rate of offshore wind power equipment is high, and its availability is 5% lower than that of onshore wind power. Therefore, it is of great significance to improve the operation and maintenance management level of offshore wind power and reduce the operation and maintenance costs.

[0003] In the existing technology, many studies have made significant progress in smart energy management systems. For example, some regions have advanced technologies and rich application experience in smart sensors, big data analysis, cloud computing, etc. At the same time, they have also demonstrated a high level in the integrated application of technologies such as the Internet of Things (IoT) and artificial intelligence (AI). In addition, the construction of new energy smart stations has also developed rapidly, and related technical research and applications have also shown a booming trend.

[0004] However, at present, the operation and maintenance management technology of offshore wind power basically copies the experience of onshore wind power. With the emergence of emerging technologies such as the Internet of Things, big data and artificial intelligence, many companies have launched intelligent operation and maintenance management systems suitable for offshore wind farms, which have been promoted and applied in the industry. However, most operation and maintenance management products are still in the research and development or trial stage and have not yet been widely used in offshore wind farms. The intelligent technologies and solutions related to the offshore wind power field are not yet mature, and the technical and management levels need to be improved urgently. Summary of the invention In order to solve the above-mentioned problems, the present invention provides an offshore wind power intelligent auxiliary data processing method.

[0005] In a first aspect, the present invention provides an offshore wind power intelligent auxiliary data processing method, which adopts the following technical solution: An offshore wind power intelligent auxiliary data processing method, comprising: The wind turbine auxiliary detection system and offshore booster station monitoring system are used for data collection respectively; The collected data are transmitted based on the station side and the station to the center side respectively; wherein, the AES algorithm is used for encryption during the transmission process; Data storage of encrypted data; The stored data is processed through the gateway system.

[0006] Furthermore, the data collection using the wind turbine auxiliary detection system and the offshore substation monitoring system includes using the wind turbine auxiliary monitoring system data to collect wind turbine fault warning data, wind turbine CMS fault warning data, wind turbine pitch bearing fault warning data, wind turbine blade status online monitoring warning data, wind turbine blade bolt loosening monitoring data, wind turbine foundation uneven settlement and tower inclination data, gearbox online oil monitoring data, wind turbine tower bolt loosening monitoring alarm data and wind farm lightning detection data.

[0007] Furthermore, the data collection using the fan auxiliary detection system and the offshore booster station monitoring system also includes using the offshore booster station monitoring system to collect data including booster station ventilation and air-conditioning monitoring data, offshore booster station electrical fire alarm data, offshore booster station comprehensive automation data, offshore booster station machine room dynamic environment monitoring data and offshore booster station structure monitoring data.

[0008] Furthermore, the data transmission of the collected data is performed based on the station side and from the station to the center side respectively, including, at the station side, converting and processing the collected data by the amplifying and filtering circuit, and then packaging it to the single-chip microcomputer through the A / D converter, and then uploading it to the IT equipment; the data is transmitted from the station side to the center side in a point-to-point form, and the station side stores the data to be reported in a designated directory. During transmission, the station side verifies, reads, encrypts, compresses, transmits, and backs up the files in the designated directory, and the center side performs corresponding reverse operations after receiving the data.

[0009] Furthermore, the AES algorithm is used for encryption during the transmission process, including expanding the input key into multiple round keys, dividing the plaintext data into 128-bit blocks and performing an XOR operation with the first round key, and then performing multiple rounds of encryption.

[0010] Furthermore, the data storage of the encrypted data includes using a distributed data storage method for data storage, and arranging the database capacity according to the data scale of the offshore wind farm, wherein the storage requirements are calculated according to the data collection frequency and data volume, and the storage architecture is designed according to the storage requirements for distributed data storage.

[0011] Furthermore, the storage architecture is designed according to storage requirements for distributed data storage, including separately setting up a data access layer, a real-time database, and a historical database, wherein the data is split into multiple fragments, each fragment is stored on a different node, the node includes a physical server or a virtual machine, and each node processes requests in parallel.

[0012] Furthermore, the stored data is processed through the gateway system, including data cleaning, data conversion and feature extraction. After data processing, the original collected data and processed statistical data are transmitted to the centralized control center or a third-party system through the data forwarding module according to the defined information model.

[0013] Furthermore, the data cleaning includes processing missing data, outliers and abnormal value data separately, using the K-nearest neighbor method to predict missing values ​​for data filling; using the Z-score method to detect abnormalities for abnormal values, and then deciding whether to delete, replace or retain abnormal values ​​based on business needs.

[0014] Furthermore, the data conversion includes normalizing the data to adapt to the requirements of the analysis tool, wherein the data is scaled to the range of [-1, 1] by a Min-Max normalization method.

[0015] In summary, the present invention has the following beneficial technical effects: 1. Reduce operation and maintenance costs: The operation and maintenance costs of offshore wind power account for a large proportion of the total project cost. By implementing intelligent auxiliary data processing methods, this cost can be significantly reduced. For example, high-precision sensors and data preprocessing technologies can capture and process key operating data in a timely manner, thereby reducing the probability of failure and the number of maintenance times.

[0016] 2. Improve equipment availability: Offshore wind power equipment has a high failure rate. Intelligent auxiliary data processing methods can predict potential equipment failures in advance and implement preventive maintenance through big data analysis and artificial intelligence technology. This will greatly improve the availability of equipment and thus improve power generation efficiency.

[0017] 3. Improve the level of operation and maintenance management: Traditional offshore wind power operation and maintenance management technologies mostly follow the experience of onshore wind power, and the level of intelligence is not high. By using emerging technologies such as the Internet of Things, big data and artificial intelligence, the level of intelligence in operation and maintenance management can be greatly improved. For example, distributed storage and efficient data transmission technology can better manage massive data, thereby improving the efficiency and response speed of operation and maintenance management.

[0018] 4. Enhanced decision support capabilities: Based on data analysis results, intelligent auxiliary data processing methods can generate optimized operation and maintenance suggestions and strategies. By displaying key information through a visual interface, operation and maintenance personnel can quickly make scientific decisions and further improve the level of operation and maintenance management. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 It is a schematic diagram of an offshore wind power intelligent auxiliary data processing method according to Example 1 of the present invention. DETAILED DESCRIPTION

[0020] The present invention is further described in detail below in conjunction with the accompanying drawings.

[0021] Example 1 Reference Figure 1 , an offshore wind power intelligent auxiliary data processing method of this embodiment includes: The wind turbine auxiliary detection system and offshore booster station monitoring system are used for data collection respectively; The collected data are transmitted based on the station side and the station to the center side respectively; wherein, the AES algorithm is used for encryption during the transmission process; Data storage of encrypted data; The stored data is processed through the gateway system.

[0022] Specifically, the following steps are included: S1. Use the wind turbine auxiliary detection system and the offshore substation monitoring system to collect data, including the use of the wind turbine auxiliary monitoring system to collect data including wind turbine fault warning data, wind turbine CMS fault warning data, wind turbine variable pitch bearing fault warning data, wind turbine blade status online monitoring warning data, wind turbine blade bolt loosening monitoring data, wind turbine foundation uneven settlement and tower inclination data, gearbox online oil monitoring data, wind turbine tower bolt loosening monitoring alarm data and wind farm lightning detection data.

[0023] The data collected by the offshore booster station monitoring system includes the booster station ventilation and air conditioning monitoring data, the offshore booster station electrical fire alarm data, the offshore booster station comprehensive automation data, the offshore booster station machine room dynamic environment monitoring data and the offshore booster station structure monitoring data.

[0024] S2. Data forwarding The data transmission protocol is separated from the transmission platform to achieve plug-in deployment of data acquisition drivers. In the system, all data acquisition and data publishing modules are pluggable, and users can flexibly load or unload various industrial protocol drivers according to their business needs. The data forwarding function module can aggregate the collected real-time data or calculated data of the equipment to generate one or more data publishing instances, which can provide data source support to third-party systems such as production management and centralized control.

[0025] S3. Data transfer In the process of data collection and transmission, the transmission process chain is relatively long, which is divided into collection, transmission, and uploading on the station side, and aggregation, reception, and storage on the center side. Ensuring the security of the data transmission solution is the key. The data transmission process mainly consists of the following two steps: 1. Data transmission on the station side: The sensors of the wind turbine auxiliary detection system and the offshore booster station detection system collect data. The collected signals are converted and processed by the amplification and filtering circuit, then packaged by the A / D converter to the microcontroller, and then uploaded to the IT equipment.

[0026] 2. Data transmission from the station side to the center side: Data is transmitted from the station side to the center side in a point-to-point manner. The station side stores the data to be reported in the designated directory. During transmission, the station side verifies, reads, encrypts, compresses, transmits, and backs up the files in the designated directory. After receiving the data, the center side performs the corresponding reverse operations.

[0027] S4. Data Encryption A complete solution for data security requirements is implemented from aspects such as package format design, name encoding, compression and decompression, data encryption, and key management. Compression and encryption are all based on data packets and are independent of each other, with independent configuration switches. The AES algorithm is used for encryption during transmission. First, the input key is expanded into multiple round keys, the plaintext data is divided into 128-bit blocks and XORed with the first round key, and then multiple rounds of encryption are performed. Each round of encryption includes the following four steps: 1. Byte Replacement: Use S-box to replace the bytes of input data.

[0028] 2. Row shift: Circularly shift the rows of the bit block to the left.

[0029] 3. Column obfuscation: Obfuscate the columns of the bit blocks through matrix multiplication operations.

[0030] 4. Round key addition: Finally, the bit block is XORed with the next key.

[0031] After multiple rounds of encryption, the last round of encryption omits the steps of byte replacement, row shift and column confusion, and directly performs an XOR operation on the bit block and the key. The encryption results of all bit blocks are used as ciphertext. The decryption process is the inverse operation of the encryption process.

[0032] S5. Data storage Data storage adopts a distributed data storage method, and the database capacity is arranged according to the data scale of the offshore wind farm, while supporting flexible capacity reduction and expansion in the future. The data forwarding terminal has the storage capacity to store at least half a year of data, and the historical database supports a data storage period of no less than 3 years, and the historical data is stored in a lossless data compression method. In addition, both the real-time database and the historical database support file backup and recovery functions.

[0033] 1. Storage capacity planning Data scale assessment: Predict the frequency and volume of data collection; assume that each wind turbine generates a certain amount of data per second (e.g., sensor data, monitoring indicators, etc.), and calculate based on actual conditions.

[0034] Storage requirement calculation: Real-time database: capacity requirement for storing data from the past six months; Historical database: supports data storage of no less than 3 years; The selection of data compression scheme is critical to space saving.

[0035] Data flow and architecture design (1) Data access layer: data collection and forwarding: Data forwarding terminals (such as edge computing devices) are used to access data, which can perform preliminary data processing (such as cleaning and aggregation) and then send the data to the main database.

[0036] (2) Real-time database: Design a streaming database (such as Apache Kafka, InfluxDB, etc.) to store real-time data, support high-frequency read and write operations, and store data from the last 6 months.

[0037] Storage architecture: Use partitioned storage, partitioned by time buckets (for example, one partition per hour) to facilitate subsequent data management.

[0038] (3) Historical database: Use a relational database (such as PostgreSQL) or a big data storage solution (such as Hadoop, ClickHouse) to save historical data.

[0039] Data compression: Implement lossless data compression methods (such as Parquet and ORC formats) to reduce the storage space occupied by historical data.

[0040] (4) Data backup and recovery: Configure regular backup strategies for real-time databases and historical databases, using a combination of incremental backup and full backup. Provide version management for backup files and support quick recovery.

[0041] Storage Tips (1) Flexible expansion and reduction: Use scalable cloud storage solutions (such as AWS S3, Google CloudStorage, etc.) to support on-demand expansion and reduction. At the database level, consider using a distributed database architecture (such as CockroachDB, Cassandra) with horizontal expansion capabilities.

[0042] (2) Data compression and storage strategy: Regularly compress historical data and use efficient storage formats to reduce storage costs. Evaluate data lifecycle management and regularly clean up expired data.

[0043] Technology Selection Real-time database: InfluxDB, Apache Kafka; Historical databases: PostgreSQL, ClickHouse, Hadoop / HDFS; Compression tools: Apache Parquet, ORC; Backup and recovery tools: Use the database's own backup tools or open source tools (such as Barman, pgBackRest); Distributed data storage method A distributed database is a database that stores data in multiple nodes, which can be physical servers or virtual machines. Each node can process requests in parallel, increasing the throughput and availability of the system.

[0044] Storage architecture design Sharding: Split the data into multiple fragments, each of which is stored on a different node. Suitable for large data sets. Reasonably define the sharding key (such as wind farm ID) and store data from different wind farms on different data nodes. Each data fragment can be copied on multiple nodes to achieve data redundancy and fault tolerance. The same data fragment is copied on 3 nodes, thereby improving the reliability and availability of the system.

[0045] Data storage selection NoSQL database: such as Cassandra, MongoDB, etc., suitable for processing unstructured and semi-structured data, supporting high concurrency and horizontal expansion.

[0046] Distributed relational databases: such as CockroachDB, Google Spanner, TiDB, etc., support ACID transactions, facilitate complex queries and data consistency management.

[0047] Database capacity layout Data generation forecast: Based on the data generated by wind farm sensors and monitoring systems, the data volume is estimated. Each wind turbine generates 1KB of data per second. If a wind farm has 100 wind turbines, 100KB will be generated per second, about 864MB per day, and about 15TB in 6 months.

[0048] Total storage requirement: Taking redundancy and backup into account, assuming three copies of storage are used, the storage requirement increases to 45TB (15TB x 3).

[0049] Node layout: Divide the data into multiple database nodes, for example, select 5 nodes as the basic nodes. Initial allocation of each node: 15TB of raw data may be allocated to 3TB per node (actually it will be adjusted dynamically as the data grows). Set the number of copies of each node data, for example, 3 copies configuration, each node stores a copy of part of the data.

[0050] Storage expansion planning When designing your solution, consider how to easily add new nodes to support future data growth. Use the automatic expansion function provided by cloud services (such as AWS and Azure) to increase storage capacity on demand.

[0051] Monitoring and optimization: Deploy monitoring tools to monitor the storage usage of each node in real time, especially indicators such as CPU, memory, I / O, and storage. When a node is close to a bottleneck, perform horizontal expansion or data migration in a timely manner.

[0052] Data lifecycle management Establish a data lifecycle strategy to manage the storage, archiving, and deletion of various types of data in an appropriate manner. Set the data retention period based on business needs. For example, real-time data is retained for 6 months, and historical data is retained for 3 years before being archived or deleted.

[0053] System architecture design Determine the architecture of the distributed database, select the appropriate database system, and design the number of nodes and regional locations (such as using geographic location distribution).

[0054] Design the data access layer to collect data from various wind turbines and equipment, and transfer the data to the distributed database through the ETL (extraction, transformation, loading) process. Perform system configuration, including copy management, sharding strategy configuration, node option configuration, etc. Deploy a monitoring system to track system performance in real time, define alarm rules, and ensure the stable operation of the database system. Develop a comprehensive backup and recovery strategy to ensure that data can be quickly restored in the event of a failure to prevent data loss.

[0055] S6. Data Governance The gateway system has data preprocessing capabilities and supports edge data processing. Data filtering is achieved by setting multiple filtering conditions, and data preprocessing is achieved through data secondary calculation, logical operation, and aggregation calculation. After data preprocessing, the original collected data and processed statistical data are transmitted to the centralized control center or third-party system through the data forwarding module according to the defined information model.

[0056] Preprocessing calculation steps: Step 1: Data cleaning It mainly processes missing data, outliers and abnormal value data. For missing data, K-nearest neighbor and decision tree methods are used to predict missing values ​​for data filling; for abnormal values, Z-score and IQR rules are used for anomaly detection, and then it is decided whether to delete, replace or retain the abnormal values ​​based on business needs.

[0057] Step 2: Data conversion This step involves recoding or standardizing the data to better suit the requirements of the analysis tool or model. The data is scaled to the range of [0, 1] or [-1, 1] through the Min-Max normalization method.

[0058] Step 3: Feature extraction Dimensionality reduction is performed through methods such as PCA and t-SNE to reduce the complexity of features, and meaningful features are extracted from the original data using methods such as statistical tests and model weights. These features may be easier to model or visualize.

[0059] The above are all preferred embodiments of the present invention, and are not intended to limit the protection scope of the present invention. Therefore, any equivalent changes made based on the structure, shape, and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. An offshore wind power intelligent auxiliary data processing method, characterized in that: include: The wind turbine auxiliary detection system and offshore booster station monitoring system are used for data collection respectively; The collected data are transmitted based on the station side and the station to the center side respectively; wherein, the AES algorithm is used for encryption during the transmission process; Data storage of encrypted data; The stored data is processed through the gateway system.

2. The offshore wind power intelligent auxiliary data processing method according to claim 1 is characterized in that: The data collection is performed using the wind turbine auxiliary detection system and the offshore substation monitoring system, including the use of the wind turbine auxiliary monitoring system data to collect wind turbine fault warning data, wind turbine CMS fault warning data, wind turbine pitch bearing fault warning data, wind turbine blade status online monitoring warning data, wind turbine blade bolt loosening monitoring data, wind turbine foundation uneven settlement and tower inclination data, gearbox online oil monitoring data, wind turbine tower bolt loosening monitoring alarm data and wind farm lightning detection data.

3. The offshore wind power intelligent auxiliary data processing method according to claim 2 is characterized in that: The data collection using the fan auxiliary detection system and the offshore booster station monitoring system also includes the use of the offshore booster station monitoring system to collect data including booster station ventilation and air conditioning monitoring data, offshore booster station electrical fire alarm data, offshore booster station comprehensive automation data, offshore booster station machine room dynamic environment monitoring data and offshore booster station structure monitoring data.

4. The offshore wind power intelligent auxiliary data processing method according to claim 3 is characterized in that: The data collected is transmitted based on the station side and from the station to the center side respectively, including that at the station side, the collected data is converted and processed by the amplification and filtering circuit, and then packaged by the single-chip microcomputer through the A / D converter, and then uploaded to the IT equipment; the data is transmitted from the station side to the center side in a point-to-point form, and the station side stores the data to be reported in a designated directory. During transmission, the station side verifies, reads, encrypts, compresses, transmits, and backs up the files in the designated directory, and the center side performs corresponding reverse operations after receiving the data.

5. The offshore wind power intelligent auxiliary data processing method according to claim 4 is characterized in that: The AES algorithm is used for encryption during the transmission process, including expanding the input key into multiple round keys, dividing the plaintext data into 128-bit blocks and performing an XOR operation with the first round key, and then performing multiple rounds of encryption.

6. The offshore wind power intelligent auxiliary data processing method according to claim 5 is characterized in that: The data storage of the encrypted data includes using a distributed data storage method to perform data storage, and arranging database capacity according to the data scale of the offshore wind farm, wherein the storage requirements are calculated according to the data collection frequency and data volume, and the storage architecture is designed according to the storage requirements to perform distributed data storage.

7. The offshore wind power intelligent auxiliary data processing method according to claim 6 is characterized in that: The storage architecture is designed according to storage requirements for distributed data storage, including separately setting up a data access layer, a real-time database, and a historical database, wherein the data is split into multiple fragments, each fragment is stored on a different node, and the node includes a physical server or a virtual machine, and each node processes requests in parallel.

8. The offshore wind power intelligent auxiliary data processing method according to claim 7 is characterized in that: The gateway system processes the stored data, including data cleaning, data conversion and feature extraction. After data processing, the original collected data and processed statistical data are transmitted to the centralized control center or a third-party system through the data forwarding module according to the defined information model.

9. The offshore wind power intelligent auxiliary data processing method according to claim 8, characterized in that: The data cleaning includes processing missing data, outliers and abnormal value data separately, using the K-nearest neighbor method to predict missing values ​​for data filling; using the Z-score method to detect abnormalities for abnormal values, and then deciding whether to delete, replace or retain abnormal values ​​based on business needs.

10. The offshore wind power intelligent auxiliary data processing method according to claim 9, characterized in that: The data conversion includes standardizing the data to adapt to the requirements of the analysis tool, wherein the data is scaled to the range of [-1, 1] by the Min-Max normalization method.

Images