Wind power plant multichannel data transmission and processing method based on 5G communication

Abstract

The invention relates to the field of communication technology and data processing, and discloses a wind power plant multichannel data transmission and processing method based on 5G communication, which comprises the following steps: collecting environment and operation data of a wind power plant through a plurality of sensors, compressing the data by adopting a compressed sensing technology, and quickly transmitting the compressed data to a data center by utilizing a 5G network, and recovering the data through a joint optimization algorithm in combination with sparse recovery and low-rank matrix recovery technologies, and finally applying the recovered data to intelligent operation, maintenance and scheduling of the wind power plant. Through the method, the data transmission quantity and the processing complexity are reduced, the data transmission speed and the recovery precision are improved, efficient technical support is provided for intelligent management, fault detection and equipment optimization of the wind power plant, meanwhile, an optimized scheduling decision is provided through data analysis, the power generation efficiency of the wind power plant is further improved, and the power generation cost of the wind power plant is reduced. Resource configuration is optimized, and operation cost is reduced.

Description

Technical Field

[0001] The present invention relates to the fields of communication technology and data processing technology, and specifically to a wind farm multi-channel data transmission and processing method based on 5G communication. Background Art

[0002] As the global demand for renewable energy continues to grow, wind energy has been widely used as a clean and sustainable energy source. As the core facilities for wind energy utilization, the scale and number of wind farms are constantly expanding, bringing higher requirements for data acquisition, transmission and processing technologies. Wind farm equipment is usually composed of a large number of sensors, which are responsible for collecting environmental parameters such as wind speed, wind direction, temperature, humidity, and the operating status data of wind turbines. As the scale of wind farms expands, the amount of data increases exponentially, and traditional data acquisition and transmission methods face increasing challenges.

[0003] Traditional data acquisition systems rely on wired communications or wireless communications based on low-speed networks. The bandwidth and transmission rate of these systems often cannot meet the needs of real-time data transmission in large-scale wind farms. Since wind farm equipment is mostly distributed in remote areas, the reliability and stability of signal transmission are required to be high, and the existing low-bandwidth, low-rate networks often cannot ensure timely and accurate data transmission. In addition, the processing capacity and efficiency of traditional data transmission methods cannot keep up with the surge in data volume, resulting in data backlogs, affecting the accuracy and timeliness of real-time decision-making.

[0004] Another technical bottleneck is data processing. In wind farms, a large amount of raw data generated by sensors needs to be processed, analyzed and stored. Existing technologies usually use traditional signal processing algorithms, such as Fourier transform, but these algorithms are complex to process and require large amounts of computation. Especially when faced with massive, time-series multi-channel data, the processing process is very time-consuming and energy-intensive. In addition, traditional processing methods have weak capabilities for identifying and repairing abnormal data, which can easily lead to misjudgment and loss of data, thus affecting the operation and maintenance and optimization decisions of wind farms.

[0005] Therefore, how to achieve efficient collection, transmission and processing of multi-channel data in wind farms, how to reduce the data transmission burden and the complexity of data processing while ensuring the real-time and accuracy of the data have become technical problems that need to be solved urgently. Summary of the invention

[0006] In view of the shortcomings of the prior art, the present invention provides a wind farm multi-channel data transmission and processing method based on 5G communication, which solves the technical problems of low transmission efficiency, high processing complexity and high energy consumption of traditional wind farm data transmission and processing methods when facing large-scale, multi-channel data.

[0007] To achieve the above objectives, the present invention is implemented through the following technical solutions: A wind farm multi-channel data transmission and processing method based on 5G communication includes the following steps: Collect multi-channel environmental data and unit operation data of wind farms through multiple sensors; Compressing the collected data and generating compressed data; Transmit compressed data to data centers via 5G communication networks; Recovering the received compressed data in the data center, the recovery step includes combining sparse recovery and low-rank matrix recovery techniques through a joint optimization algorithm; The recovered data is used for intelligent operation, maintenance and scheduling of wind farms, including fault detection, equipment analysis and optimization decisions.

[0008] Preferably, the method further includes the step of preprocessing the multi-channel environmental data and the unit operation data, specifically: Performing denoising processing on the collected raw data; Standardize the denoised data; Fill missing values ​​in the processed data; Perform timing alignment on the data.

[0009] Preferably, the step of compressing the collected data and generating compressed data comprises: Design a random measurement matrix to map the standardized data matrix to a low-dimensional space, wherein the measurement matrix is ​​a sparse matrix; The standardized data is linearly transformed through the measurement matrix to obtain the compressed data matrix, and the compression accuracy is controlled by an error tolerance.

[0010] Preferably, the step of restoring the received compressed data in the data center includes: Recover compressed data through sparse recovery technology to minimize norm to enforce sparsity constraints; Using low-rank matrix recovery technology, the low-rank part is restored by minimizing the nuclear norm of the matrix; Through the joint optimization model, weighted optimization is performed between sparse recovery and low-rank matrix recovery to obtain the restored complete data.

[0011] Preferably, in the process of restoring the compressed data, a weighted sparsity recovery technology is further adopted, specifically including: Assign a weight matrix to each sensor, each element of which represents the sparsity of the sensor data; By weighting The norm minimization algorithm recovers the sparse parts of different sensors and finally obtains the weighted recovered data.

[0012] Preferably, a multi-level optimization framework is used in the process of restoring the compressed data, comprising the following steps: The wind farm data in the recovery process is divided into multiple sub-matrices, each sub-matrix represents different types of data; For each data submatrix, sparse recovery and low-rank matrix recovery methods are applied for local optimization; All sub-matrix recovery results are combined to obtain complete wind farm data.

[0013] Preferably, the step of transmitting the compressed data to the data center via the 5G communication network includes: A dynamic bandwidth allocation mechanism is adopted to adjust the data compression ratio according to network bandwidth and data transmission requirements.

[0014] Preferably, the accuracy of the recovery step is evaluated by the following indicators: The accuracy of the restoration is evaluated by calculating the mean square error between the restored data and the original data; Use peak signal-to-noise ratio as a measure of data recovery quality The present invention also provides a wind farm multi-channel data transmission and processing device based on 5G communication, including: Data acquisition module, used to collect multi-channel environmental data and unit operation data of wind farms; A compression module is used to compress the collected data to generate compressed data; A communication module, used to transmit compressed data to a data center via a 5G communication network; The data recovery module is used to receive and recover the compressed data. The recovery process uses a joint optimization algorithm combined with sparse recovery and low-rank matrix recovery; The application module is used to use the recovered data for intelligent operation, maintenance and scheduling of wind farms.

[0015] The present invention also provides a computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the above method is implemented.

[0016] The present invention also provides a storage medium on which a computer program is stored. When the computer program is executed by a processor, the method described above is implemented.

[0017] The present invention provides a wind farm multi-channel data transmission and processing method based on 5G communication. It has the following beneficial effects: 1. By combining compressed sensing technology and 5G communication network, the present invention effectively reduces the amount of data and improves the transmission speed. Compressed sensing technology can compress data without losing key information, reducing the pressure on transmission bandwidth, especially in large-scale wind farms. The low latency and high bandwidth characteristics of 5G networks ensure efficient and real-time transmission of data. This allows the various sensor data of the wind farm to be quickly transmitted to the data center, meeting the needs of large-scale, multi-channel data real-time processing.

[0018] 2. The present invention achieves effective dimensionality reduction and optimization in the data recovery process by adopting sparse recovery and low-rank matrix recovery technology. In particular, by combining the joint optimization algorithm with the sparse recovery and low-rank matrix recovery methods, the complexity and amount of calculation of data processing are significantly reduced. This not only improves the accuracy of recovery, but also effectively reduces the energy consumption in the data processing process, meeting the requirements of low power consumption and efficient calculation for intelligent management of wind farms.

[0019] 3. The restored data provides accurate information support for the intelligent operation and maintenance and dispatching of wind farms. By combining fault detection, equipment analysis and optimized decision-making functions, it can monitor the operating status of wind turbines in real time, detect potential faults in time and issue early warnings, thereby improving the reliability of wind farms and the service life of equipment. At the same time, through data analysis, it provides optimized dispatching decisions, further improves the power generation efficiency of wind farms, optimizes resource allocation, and reduces operating costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 It is a schematic diagram of the method flow of the present invention; Figure 2 It is a schematic diagram of the structure of the device of the present invention; Figure 3 It is a schematic diagram of the computer device structure of the present invention.

[0021] Among them, 100, data acquisition module; 200, compression module; 300, communication module; 400, data recovery module; 500, application module; 40, computer equipment; 41, processor; 42, memory; 43, storage medium. DETAILED DESCRIPTION

[0022] The following will be combined with the drawings in the specification of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0023] Please see attached Figure 1The wind farm multi-channel data transmission and processing method based on 5G communication of the present invention can effectively improve the intelligent operation and maintenance level of wind farms and solve the bottleneck problems of data collection, transmission and processing in large-scale wind farms. By utilizing the high speed, low latency and large number of connections of 5G communication technology, and combining compressed sensing technology with data recovery algorithm, efficient collection, compression, transmission and recovery of wind farm multi-channel data are achieved. The following is a detailed description of the specific implementation methods of the present invention.

[0024] As shown in FIG. 1 , the wind farm multi-channel data transmission and processing method based on 5G communication of the present invention may include the following steps: S1. Collect multi-channel environmental data and unit operation data of the wind farm through multiple sensors; S2. compressing the collected data and generating compressed data; S3, transmit the compressed data to the data center via the 5G communication network; S4, recovering the received compressed data in the data center, the recovery step comprising combining sparse recovery and low-rank matrix recovery techniques through a joint optimization algorithm; S5. Use the recovered data for intelligent operation, maintenance and dispatching of wind farms, including fault detection, equipment analysis and optimization decision-making.

[0025] In this embodiment, in step S1, the wind farm collects multi-channel environmental data and unit operation data in real time through multiple sensors to ensure the comprehensiveness and accuracy of the required data. The sensors mainly include wind speed sensors, wind direction sensors, temperature sensors, humidity sensors, air pressure sensors, and unit operation status sensors (such as speed, vibration, power output, etc.). These data are not only used to monitor environmental conditions and equipment conditions in real time, but also provide a basis for subsequent data processing and analysis.

[0026] As an option, the acquisition frequency of the sensor is determined according to the specific monitoring requirements and the performance characteristics of the sensor. Generally, the acquisition frequency is between 1 second and 10 seconds. For some sensors (such as environmental monitoring sensors), the acquisition frequency may be lower than the acquisition frequency of the unit operation data, but data synchronization is still very important.

[0027] It should be noted that in wind farms, the multi-channel characteristics of data acquisition make data processing and transmission more complicated. Multi-channel data usually involves multiple dimensions, the amount of data is huge, and the sampling timestamp and measurement unit of each channel may be different. Therefore, how to effectively and accurately collect these multi-dimensional data and prepare for subsequent data compression and transmission has become one of the key technologies in the present invention.

[0028] Specifically, the data collection steps include the following aspects: Environmental data collection: including real-time monitoring of environmental factors such as wind speed, wind direction, temperature, humidity, and air pressure. These data are usually collected by environmental monitoring sensors installed at different locations in the wind farm. Each environmental sensor can provide different measurement accuracy and sampling period according to specific needs. For example, wind speed and wind direction sensors can usually provide high real-time performance and accuracy, ensuring that changes in wind resources can be fully reflected.

[0029] Unit operation data collection: including various operating parameters of wind turbines, such as speed, power output, vibration signal, current, voltage, etc. These data are collected in real time through sensors installed inside the wind turbine (such as vibration sensors, speed sensors, power sensors). These data are crucial for unit operation status analysis and fault prediction. Specifically, the frequency of unit data collection is usually high, at least once per second, so as to accurately capture the dynamic changes of the unit.

[0030] Data synchronization and timestamp management: The data collected by various sensors in wind farms are often distributed, so data synchronization must be ensured through reasonable timestamp management. In order to ensure the time consistency of multi-channel data, each data point must be accurately timestamped during data collection. These timestamps will serve as the basis for subsequent data processing to facilitate time alignment, data fusion and analysis.

[0031] In one embodiment, data collection is performed by a centralized data collector. The data collector receives real-time data from various sensors and temporarily stores it in a local cache. In the pre-processing phase after collection, these data will first undergo denoising, standardization, missing value filling and other operations to improve data quality and prepare for subsequent processing.

[0032] It should be noted that the selection and layout of sensors have a direct impact on the accuracy and comprehensiveness of the data. In the present invention, by rationally configuring different types of sensors, it is possible to ensure that the multi-channel data collection of the wind farm has comprehensive coverage and can provide high-precision data input.

[0033] After data collection, the obtained raw data usually needs to be preprocessed to ensure its quality and consistency. In the embodiment of the present invention, the data preprocessing step mainly includes four sub-steps: denoising, standardization, missing value filling and time series alignment. Each step has an important impact on the quality of the data and subsequent processing.

[0034] De-noising: Since sensors in wind farms are often affected by environmental noise, electromagnetic interference and other factors, there may be noise in the collected data. In order to improve the quality of the data, denoising is required. Common denoising methods include Kalman filtering and wavelet transform.

[0035] Kalman filter method: As a recursive algorithm, Kalman filter can effectively filter out noise interference. Its basic idea is to minimize the estimation error based on the system model and noise model, combined with the observed data, so as to provide the best estimate of the real data.

[0036] Wavelet transform method: Wavelet transform is a very effective signal denoising method, especially suitable for signals with mutation or non-stationary characteristics. Through wavelet transform, the signal can be decomposed into components of different scales, and the low-frequency components can be used to remove high-frequency noise, thereby effectively removing noise from the data.

[0037] Standardization: To ensure the consistency of the dimensions and value range of data collected by different sensors, the data usually needs to be standardized. Standardization is usually achieved by subtracting the mean of each data point and dividing it by the standard deviation, so that the data of each data channel is in a unified standard range.

[0038] Missing value filling: In practical applications, sensors may be missing some data due to failures or communication problems. In order to ensure data integrity, missing data needs to be filled. Common filling methods include neighborhood-based data interpolation and linear regression model filling.

[0039] Interpolation: Interpolation estimates missing values ​​using values ​​between known data points. For example, linear interpolation estimates missing data by connecting previous and next data points to generate a straight line.

[0040] Regression Models: In some cases, data from other sensors can be used to build regression models to predict missing values ​​based on existing observations.

[0041] Time alignment: Since various sensors in a wind farm may have different acquisition timestamps, it is necessary to time-align these data. The purpose of time alignment is to ensure that data from different sensors can be compared and fused within the same time window during data analysis and processing.

[0042] Interpolation method: For sampling data with different timestamps, the time step of the data is usually adjusted through interpolation methods (such as linear interpolation or spline interpolation) to ensure that all data can be aligned on a unified time axis.

[0043] It should be pointed out that the effectiveness of the data preprocessing step directly affects the subsequent data recovery and analysis results. In the present invention, the denoising, standardization, missing value filling and time series alignment methods are used to maximize the data quality and ensure the accuracy and effectiveness of the subsequent compressed sensing and data recovery process.

[0044] In this embodiment, the environmental data and unit operation data of the wind farm are collected through multiple sensors. After accurate preprocessing, the obtained multi-channel data will provide high-quality input data for subsequent compression, transmission, recovery and other steps.

[0045] In this embodiment, the purpose of step S2 is to compress the collected multi-channel data through compressed sensing technology to reduce the bandwidth required for data transmission, reduce the complexity of data processing, and retain the key information of the data as much as possible. The compressed data will be used for subsequent data transmission and recovery.

[0046] Specifically, the compression process mainly designs a random measurement matrix and maps the standardized data matrix to a low-dimensional space. This process uses the linear transformation method in compressed sensing theory to reduce the dimensionality of the original data and effectively compress the data while ensuring the quality of data recovery.

[0047] First, design a sparse measurement matrix , the matrix is ​​a The matrix of , which means mapping the original data from high-dimensional space to low-dimensional space. As an option, Can be set to the original data dimension A small proportion value (for example, 10%) of . Through this measurement matrix, the original high-dimensional data will be projected into a lower-dimensional space, thereby achieving data compression.

[0048] Specifically, assuming the original data matrix is , whose dimensions are (in is the dimension of the data, is the time step or the number of data samples), then the compressed data matrix It can be expressed by the following formula: in, is the compressed data matrix with dimension , is the designed measurement matrix with dimension . Through matrix multiplication , mapping the original data to a low-dimensional space, thereby achieving data compression.

[0049] It should be noted that when designing the measurement matrix When performing a sparse matrix, a sparse matrix is ​​usually used, which means that most elements in the matrix are zero, and non-zero elements are distributed in a few positions of the matrix. Sparse matrices have lower storage requirements and can effectively reduce computational complexity while ensuring the effectiveness of data compression.

[0050] In the compression process, in order to ensure the quality of data recovery, the compression accuracy needs to be controlled. In one possible implementation, an error tolerance can be introduced. To control the compression accuracy. This tolerance is used to balance the relationship between data compression rate and recovery accuracy during the compression process. Specifically, the compression process selects a suitable measurement matrix through an optimization algorithm. , so that the error between the compressed data and the original data remains within an acceptable range.

[0051] For each compressed data matrix , the recovery error can be calculated by the following formula: in, is the measurement matrix The pseudo-reversal, yes The norm indicates the size of the error. This formula ensures that the difference between the compressed data and the original data during the recovery process will not exceed the preset error tolerance. .

[0052] Specifically, the key to compressed sensing lies in the compressed data matrix Can the original data be effectively reconstructed through subsequent recovery algorithms? In order to achieve higher recovery accuracy, it is necessary to select a suitable error tolerance , which is usually related to the compression ratio Inversely proportional, that is, the higher the compression rate, the greater the error tolerance needs to be.

[0053] In the implementation process, choosing a suitable measurement matrix is ​​crucial for the compression effect. Generally, the design of the measurement matrix can be based on random sampling, Hadamard matrix, or other matrix structures with good mathematical properties. In an embodiment of the present invention, the measurement matrix adopts a random measurement matrix design method, which can ensure low computational complexity and has good recovery performance.

[0054] In one possible implementation, the random measurement matrix The elements of are randomly generated using independent and identically distributed Gaussian or Bernoulli distributions. For example, the matrix Each element of can be generated as follows: in, is from a standard normal distribution A random variable, is the dimension of the data matrix. The measurement matrix generated in this way has good compressed sensing performance and can effectively retain the main features of the data in a low-dimensional space.

[0055] The final result of the compression process is a low-dimensional compressed data matrix , the matrix contains most of the important information of the original data, and its data volume is greatly reduced. The compressed data will be further transmitted to the data center through the 5G communication network. During the transmission process, since the amount of data is greatly reduced, the bandwidth consumption can be significantly reduced and the data transmission efficiency can be improved.

[0056] It is important to understand that the compressed data matrix It does not simply "discard" information, but retains the most important features and information in the data through reasonable dimensionality reduction and compression strategies. This ensures the effectiveness of data transmission and high accuracy of data recovery even under low bandwidth conditions.

[0057] By designing a sparse measurement matrix and using compressed sensing technology to effectively compress the multi-channel data collected by the wind farm, not only can the amount of data be significantly reduced, but also the data transmission efficiency can be improved while ensuring the accuracy of data recovery. In this embodiment, by optimizing the measurement matrix and controlling the compression accuracy, efficient data compression can be achieved, and the key information of the data can be retained to the maximum extent in the subsequent recovery steps. The compressed data will be efficiently transmitted under the 5G communication network, providing high-quality data support for the intelligent operation and maintenance and scheduling of the wind farm.

[0058] In this embodiment, step S3 mainly transmits the compressed data from the wind farm to the data center through the 5G communication network. The core goal of this step is to achieve efficient data transmission and maximize the use of network bandwidth resources during the transmission process. With the support of 5G communication technology, it can provide the advantages of high speed, low latency and large-scale device connection. The present invention makes full use of the characteristics of the 5G communication network to improve data transmission efficiency.

[0059] Specifically, data transmission is regulated by adopting a dynamic bandwidth allocation mechanism. The core of this mechanism is to dynamically adjust the data compression ratio according to the actual network bandwidth and the demand for data transmission. This method can optimize transmission efficiency under different network conditions and ensure that data can be transmitted to the data center with the lowest delay and high reliability under limited bandwidth conditions.

[0060] The dynamic bandwidth allocation mechanism in this embodiment first performs real-time monitoring and calculation based on the current network bandwidth status, communication load, and the size of the transmitted data, thereby determining the optimal ratio of data compression. The compression ratio may be different for different network bandwidths and transmission conditions. When the bandwidth is high, the system may choose a lower compression ratio to ensure the integrity and recovery accuracy of data transmission; when the bandwidth is low, the system may increase the compression ratio and reduce the amount of transmitted data to adapt to the bandwidth limitations of the network.

[0061] Specifically, the dynamic bandwidth allocation mechanism can be implemented through the following steps: Network bandwidth monitoring: By monitoring the bandwidth status of the 5G communication network in real time, the currently available bandwidth can be obtained. Bandwidth monitoring can be achieved through spectrum analysis, bandwidth detection and other methods, and can reflect the actual transmission capacity of the network in real time.

[0062] Transmission requirements analysis: Evaluate the data transmission requirements based on the priority of the data and the urgency of transmission. For example, real-time data transmission for a wind farm may require low latency and high bandwidth, while non-real-time data can be transmitted with lower bandwidth.

[0063] Compression ratio calculation: Based on the network bandwidth and data transmission requirements, an algorithm is used to calculate the best compression ratio. The choice of compression ratio is not only limited by the network bandwidth, but also takes into account the accuracy requirements of data recovery. For some data that does not require extremely high accuracy, a higher compression ratio can be selected to save bandwidth; for data that requires higher accuracy, the compression ratio is correspondingly lower.

[0064] Dynamic adjustment of bandwidth and compression ratio: During data transmission, the system will dynamically adjust the compression ratio according to the real-time changes in network bandwidth and transmission requirements. If the network bandwidth changes, the system will re-evaluate the compression ratio and change the compression strategy if necessary. By dynamically adjusting the compression ratio, it is possible to minimize bandwidth waste while ensuring transmission efficiency.

[0065] For example, when the network bandwidth is insufficient, the compression ratio can be increased to a higher value to reduce the amount of data that needs to be transmitted and avoid network congestion. Conversely, when the bandwidth is high, a lower compression ratio can be selected to ensure data recovery accuracy.

[0066] During the data transmission process, in order to ensure the stability and reliability of the transmission, the transmitted data will be divided into multiple data packets for processing. Each data packet contains a part of the compressed data, and each data packet has an independent checksum for error detection and repair during the transmission process. Data packet management during the transmission process includes the following aspects: Data packet division: Compress the data matrix Divide into multiple small data packets. Each data packet contains a part of the compressed data matrix and related metadata (for example, data packet number, checksum information, etc.). In this way, the packet loss problem during transmission can be effectively avoided, and each data packet can be transmitted and recovered independently.

[0067] Error detection and verification: To ensure the reliability of data transmission, each data packet contains a check code (such as a CRC check code or a hash value) to detect errors in the data during transmission. When the receiving data center receives the data packet, it will first check it and request retransmission of the corresponding data packet if an error is found.

[0068] Retransmission mechanism: During the transmission process, if a data packet is lost or erroneous, the receiver will request the sender to retransmit it based on the checksum. In this way, the system can improve the reliability of data transmission without affecting the overall transmission efficiency.

[0069] It should be noted that the low latency and high bandwidth characteristics of the 5G communication network provide very important technical support for the present invention. Traditional wireless communication networks usually have high latency and bandwidth limitations, which make it difficult to meet the needs of wind farms for real-time data transmission. 5G communication technology provides lower latency and greater bandwidth, which can meet the real-time transmission needs of large-scale data. In this embodiment, the low latency characteristics of the 5G network are particularly important because the intelligent operation and maintenance and scheduling of wind farms often require rapid processing of real-time data.

[0070] In this embodiment, in order to further ensure the reliability of data transmission, the redundancy mechanism and multi-path transmission technology of the 5G network are adopted. Through multi-path transmission, data can be transmitted simultaneously through multiple paths to avoid network congestion or failure of a single path affecting data transmission. This multi-path transmission method can ensure that data can be transmitted to the data center stably and efficiently in a complex network environment.

[0071] Through the dynamic bandwidth allocation mechanism of the 5G communication network, this embodiment can intelligently adjust the data compression ratio according to the network bandwidth and data transmission requirements to ensure the balance between efficiency and accuracy during data transmission. With this mechanism, it is possible to optimize the transmission strategy under different network conditions, reduce the bandwidth occupancy of data transmission, improve transmission efficiency, and ensure data recovery accuracy. Through real-time bandwidth monitoring, transmission demand analysis, and compression ratio calculation, the system can automatically adjust the compression ratio to ensure that data transmission tasks can be completed efficiently and stably under different network environments.

[0072] In this embodiment, step S4 is to recover the compressed data transmitted to the data center through the 5G communication network to ensure that the multi-channel data of the wind farm can accurately reflect its original information. Since the data has been compressed during the transmission process, in order to ensure the validity and accuracy of the data, it must be restored through a recovery algorithm. A series of technologies are used in the recovery process, including sparse recovery, low-rank matrix recovery and joint optimization model methods to optimize the recovery accuracy.

[0073] The recovery process is divided into multiple stages, and through gradual optimization and refined calculation, the recovered data is ensured to be as close to the original data as possible. Specifically, the recovery techniques used in this step include sparse recovery technology, low-rank matrix recovery technology, weighted sparsity recovery technology, and multi-level optimization framework. These technologies work together through different optimization strategies to obtain high-precision recovery results.

[0074] Firstly, sparse recovery technology and low-rank matrix recovery technology are used to process different parts of the data respectively to improve the recovery accuracy.

[0075] Sparse recovery: In some data channels, the data collected by the sensor may be sparse, that is, most of the data values ​​are zero or close to zero. For such sparse data, the sparsity constraint can be strengthened by minimizing the norm. The sparsity of the data will be considered during the recovery process, thereby improving the efficiency and accuracy of the recovery. Specifically, sparse recovery can be expressed as the following optimization problem: in, represents the sparse data after recovery, is the measurement matrix during compression, is the received compressed data, represents the recovery error. By minimizing , even if the data presents sparse characteristics, the system can effectively recover the non-zero part of the data.

[0076] Low-rank matrix recovery: On the other hand, for some data containing dense information, low-rank matrix recovery technology is more applicable. By minimizing the nuclear norm of the matrix (i.e., the rank of the matrix), this technology can recover the low-rank part contained in the data, that is, those data with strong intrinsic structure. The optimization problem of low-rank matrix recovery can be expressed as: in, is the restored low-rank matrix, is the measurement matrix, is the received compressed data, is the recovery error term, the nuclear norm Is the rank measure of the matrix, and minimizing the nuclear norm helps to recover the low-rank information hidden in the data.

[0077] In order to find the best balance between sparse recovery and low-rank matrix recovery, a joint optimization model is introduced in this embodiment. This model combines the advantages of sparse recovery and low-rank matrix recovery through weighted optimization to further improve the recovery accuracy.

[0078] The goal of the joint optimization model is to minimize the recovery error and weight between sparse recovery and low-rank matrix recovery to achieve a balance between accuracy and efficiency. Specifically, the joint optimization problem can be expressed as: in, and are weighted coefficients, which control the weights of low-rank recovery and sparse recovery respectively. This optimization problem can flexibly adjust the sparsity and low-rank constraints at different stages of the recovery process, so as to better adapt to the different characteristics of wind farm data.

[0079] In order to further improve the recovery accuracy, the present embodiment also adopts the weighted sparsity recovery technology. This technology adjusts the recovery accuracy of different sensor data by assigning a weight matrix to each sensor. The elements of each weight matrix represent the sparsity of the sensor data, thereby performing differentiated processing on the recovery process of different sensors.

[0080] Specifically, in the weighted sparsity recovery technique, a weight matrix is ​​first assigned to each sensor ,in is the sensor number, Each element of represents the sparsity of the sensor data. Then, the sparse parts of different sensors are restored through the weighted norm minimization algorithm, and finally the weighted restored data is obtained. This process can be expressed as the following optimization problem: in, It is The data recovery results of the sensors are is the weight matrix of the sensor, and the optimization goal is to recover the data by minimizing the weighted norm.

[0081] In the process of restoring the compressed data, this embodiment also adopts a multi-level optimization framework, decomposing the restoration process into multiple sub-stages, and performing local optimization on different types of wind farm data respectively.

[0082] First, the wind farm data in the recovery process is divided into multiple sub-matrices, each of which represents different types of data in the wind farm (for example, wind speed, temperature, humidity, etc.). Then, for each sub-matrix, sparse recovery and low-rank matrix recovery methods are applied for local optimization to improve the recovery accuracy. The optimization process of each sub-matrix can be processed by different algorithms, and the most appropriate recovery method is selected according to the characteristics of the data.

[0083] Finally, all sub-matrix recovery results are combined to obtain complete wind farm data. This multi-level optimization framework can select appropriate recovery methods according to different data characteristics, thereby ensuring recovery accuracy and efficiency.

[0084] In this embodiment, the accuracy of the recovery step is evaluated by the following indicators: Mean square error (MSE): The accuracy of the recovery is evaluated by calculating the mean square error between the recovered data and the original data. The mean square error is defined as: in, is the restored data value, is the original data value, is the total number of data points. A smaller mean square error indicates a higher recovery accuracy.

[0085] Peak signal-to-noise ratio (PSNR): The peak signal-to-noise ratio is used as a measure of data recovery quality. The peak signal-to-noise ratio is defined as: in, is the maximum value in the recovered data. A higher peak signal-to-noise ratio indicates higher quality of data recovery.

[0086] The data recovery process in this embodiment uses a variety of advanced optimization techniques, including sparse recovery, low-rank matrix recovery, weighted sparsity recovery, and a multi-level optimization framework. Through the combined use of these techniques, important information in the compressed data can be recovered to the greatest extent and the accuracy of the recovery can be improved. The strategy of weighted sparse recovery and low-rank matrix recovery using a joint optimization model can be flexibly adjusted under different recovery requirements to ensure the accuracy and reliability of the data. Through these technologies, the present invention can achieve efficient and accurate recovery of multi-channel data in wind farms, providing a reliable foundation for subsequent data processing and analysis.

[0087] In this embodiment, step S5 is to use the restored data to perform intelligent operation and maintenance and scheduling of the wind farm. This process combines the restored environmental data and unit status data with the wind farm management system to achieve key functions such as fault detection, equipment analysis and optimization decision-making. By analyzing the restored data, the operating status of the wind farm can be fully monitored and scheduled in real time, the working efficiency of the wind turbines can be optimized, and the overall performance and reliability of the wind farm can be improved.

[0088] Specifically, the restored data includes environmental parameters (such as wind speed, wind direction, temperature, humidity, etc.) and unit operating status data (such as speed, power output, vibration, etc.) collected by multiple sensors. Through the comprehensive analysis of these data, targeted optimization strategies can be provided. The following describes in detail the specific application scenarios and technical implementations of intelligent operation and maintenance and scheduling.

[0089] Fault detection is a key task in the intelligent operation and maintenance of wind farms. It aims to detect equipment failures in a timely manner and issue early warnings through real-time monitoring of equipment status data. The recovered data provides a reliable source of information for fault detection, especially in the operating status data of the unit, where certain abnormal patterns may indicate potential equipment failures.

[0090] In some embodiments, data mining and machine learning algorithms can be used to perform pattern recognition on the restored wind farm data. For example, clustering algorithms (such as K-means) or support vector machines (SVM) can be used to analyze the vibration data, temperature changes, power fluctuations and other characteristics of the wind turbines to identify potential failure modes.

[0091] For example, if the vibration acceleration of the unit exceeds a certain set threshold, it may indicate blade wear or bearing damage. At this time, the system will output an alarm signal based on the recovered data and promptly notify the operation and maintenance personnel to conduct on-site inspections.

[0092] During the fault detection process, the health index of the device can be calculated using the following formula: in, For the Item sensor data, For the restored data, is the weight coefficient of each data item, Is the total number of data items. By calculating the health index of the device, you can determine whether the device is in normal operation.

[0093] Equipment analysis refers to a comprehensive analysis of the operating data of each wind turbine in a wind farm to evaluate its performance and provide an optimized operation plan. The goal of equipment analysis is to find out the operating bottleneck of the equipment and formulate a more reasonable scheduling strategy by deeply mining the recovered data.

[0094] In some embodiments, equipment analysis can be modeled based on the relationship between the power output of the restored wind turbine and environmental parameters. Through multiple regression analysis or neural network methods, the working efficiency of the wind turbine under different environmental conditions can be evaluated to discover the potential and optimization space of the equipment.

[0095] Specifically, assuming that there is a certain nonlinear relationship between the recovered wind speed data and the unit power output, this relationship can be fitted by establishing a regression model: in, Indicates the power output, is the restored wind speed data, is the fitting function, is the error term. Through the optimization model, the optimal power output of wind turbines under different wind speed conditions can be predicted, and optimization decision support can be provided for the dispatching system.

[0096] In the intelligent dispatching process of wind farms, the optimization decision aims to formulate the optimal unit dispatching strategy based on the operating data of wind turbines and the distribution of wind energy resources. This process requires full use of the recovered data to provide a scientific basis for the operating efficiency of wind farms.

[0097] In some embodiments, an optimal dispatch model for wind farms can be established to calculate the optimal start-up and shutdown decisions of wind farms in combination with weather forecasts and grid demand. The model can minimize the overall operating cost of wind farms or improve power generation efficiency based on restored data through optimization algorithms such as linear programming (LP) or integer programming (IP).

[0098] For example, suppose there are The goal is to maximize the total power output of the wind farm. The constraints include wind speed, unit status, equipment failure and other factors. The optimization model can be expressed as the following mathematical formula: in, For the The power output of each unit, is the restored wind speed data, is the maximum power output of the unit, is the on / off status of the unit (1 means on, 0 means off). The constraints ensure the optimal operation of the unit under given wind speed conditions.

[0099] In order to further improve the accuracy and flexibility of dispatching decisions, an intelligent feedback mechanism is introduced in this embodiment. By continuously monitoring and analyzing the real-time data of the wind farm, the system can continuously adjust the dispatching strategy according to the operating conditions to adapt to the dynamically changing environment.

[0100] In one possible implementation, the intelligent dispatching system uses an adaptive algorithm combined with machine learning technology to continuously optimize the dispatching strategy based on historical data and real-time data. For example, the system can use reinforcement learning (RL) algorithms to learn how to adjust the start and stop strategies of the units under different wind speeds and load conditions, thereby achieving dynamic optimization of wind farm dispatching.

[0101] In this embodiment, step S5 realizes intelligent operation and maintenance and dispatch of the wind farm by combining the restored wind farm data. Through functions such as fault detection, equipment analysis and optimization decision-making, the system can effectively improve the operating efficiency of wind turbines, reduce downtime, and improve the overall reliability of wind farms. Through optimization models and intelligent feedback mechanisms, the system can dynamically adjust the operating strategy according to real-time data and environmental changes to achieve intelligent management of wind farms.

[0102] In general, the present invention realizes the efficient collection, transmission and processing of multi-channel environmental data and unit operation data of wind farms by combining compressed sensing technology, fast Fourier transform (FFT) and 5G communication network. The method compresses and transmits the data collected by sensors, and restores high-precision complete data in the data center through a joint optimization algorithm. The restored data can be used for intelligent operation and maintenance and scheduling of wind farms, including fault detection, equipment analysis and optimization decision-making, thereby improving the operating efficiency and reliability of wind farms.

[0103] The wind farm multi-channel data transmission and processing device based on 5G communication described below and the wind farm multi-channel data transmission and processing method based on 5G communication described above can be referenced to each other.

[0104] Please see attached Figure 2 The present invention also provides a wind farm multi-channel data transmission and processing device based on 5G communication, including: The data acquisition module 100 is used to collect multi-channel environmental data and unit operation data of the wind farm; The compression module 200 is used to compress the collected data to generate compressed data; The communication module 300 is used to transmit the compressed data to the data center via the 5G communication network; A data recovery module 400 is used to receive and recover compressed data, and a joint optimization algorithm is used in the recovery process to combine sparse recovery and low-rank matrix recovery; The application module 500 is used to use the restored data for intelligent operation, maintenance and scheduling of the wind farm.

[0105] The device of this embodiment can be used to execute the above method embodiment, and its principles and technical effects are similar, which will not be repeated here.

[0106] Please see attached Figure 3 The present invention further provides a computer device 40, comprising: a processor 41 and a memory 42, wherein the memory 42 stores a computer program executable by the processor, and when the computer program is executed by the processor, the above method is executed.

[0107] The present invention further provides a storage medium 43 on which a computer program is stored. When the computer program is run by the processor 41, the above method is executed.

[0108] Among them, the storage medium 43 can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable red-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, disk or optical disk.

[0109] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A wind farm multi-channel data transmission and processing method based on 5G communication, characterized in that: The following steps are involved: Collect multi-channel environmental data and unit operation data of wind farms through multiple sensors; Compressing the collected data and generating compressed data; Transmit compressed data to data centers via 5G communication networks; Recovering the received compressed data in the data center, the recovery step includes combining sparse recovery and low-rank matrix recovery techniques through a joint optimization algorithm; The recovered data is used for intelligent operation, maintenance and scheduling of wind farms, including fault detection, equipment analysis and optimization decisions.

2. The wind farm multi-channel data transmission and processing method based on 5G communication according to claim 1 is characterized in that: The method also includes the steps of preprocessing the multi-channel environmental data and the unit operation data, specifically: Performing denoising processing on the collected raw data; Standardize the denoised data; Fill missing values ​​in the processed data; Perform timing alignment on the data.

3. The wind farm multi-channel data transmission and processing method based on 5G communication according to claim 1 is characterized in that: The step of compressing the collected data and generating compressed data comprises: Design a random measurement matrix to map the standardized data matrix to a low-dimensional space, wherein the measurement matrix is ​​a sparse matrix; The standardized data is linearly transformed through the measurement matrix to obtain the compressed data matrix, and the compression accuracy is controlled by an error tolerance.

4. The wind farm multi-channel data transmission and processing method based on 5G communication according to claim 1 is characterized in that: The step of restoring the received compressed data in the data center includes: Recover compressed data through sparse recovery technology to minimize norm to enforce sparsity constraints; Using low-rank matrix recovery technology, the low-rank part is restored by minimizing the nuclear norm of the matrix; Through the joint optimization model, weighted optimization is performed between sparse recovery and low-rank matrix recovery to obtain the restored complete data.

5. The wind farm multi-channel data transmission and processing method based on 5G communication according to claim 4 is characterized in that: In the process of restoring the compressed data, a weighted sparsity recovery technique is further adopted, specifically including: Assign a weight matrix to each sensor, each element of which represents the sparsity of the sensor data; By weighting The norm minimization algorithm recovers the sparse parts of different sensors and finally obtains the weighted recovered data.

6. The wind farm multi-channel data transmission and processing method based on 5G communication according to claim 1 is characterized in that: A multi-level optimization framework is used in the process of restoring the compressed data, including the following steps: The wind farm data in the recovery process is divided into multiple sub-matrices, each sub-matrix represents different types of data; For each data submatrix, sparse recovery and low-rank matrix recovery methods are applied for local optimization; All sub-matrix recovery results are combined to obtain complete wind farm data.

7. The wind farm multi-channel data transmission and processing method based on 5G communication according to claim 1 is characterized in that: The step of transmitting the compressed data to the data center via the 5G communication network includes: A dynamic bandwidth allocation mechanism is adopted to adjust the data compression ratio according to network bandwidth and data transmission requirements.

8. The wind farm multi-channel data transmission and processing method based on 5G communication according to claim 1 is characterized in that: The accuracy of the recovery step is evaluated using the following metrics: The accuracy of the restoration is evaluated by calculating the mean square error between the restored data and the original data; The peak signal-to-noise ratio is used as a measure of the quality of data recovery.

9. A wind farm multi-channel data transmission and processing device based on 5G communication, applied to the method according to any one of claims 1 to 8, characterized in that: include: Data acquisition module, used to collect multi-channel environmental data and unit operation data of wind farms; A compression module is used to compress the collected data to generate compressed data; A communication module, used to transmit compressed data to a data center via a 5G communication network; The data recovery module is used to receive and recover the compressed data. The recovery process uses a joint optimization algorithm combined with sparse recovery and low-rank matrix recovery; The application module is used to use the recovered data for intelligent operation, maintenance and scheduling of wind farms.

10. A storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method according to any one of claims 1 to 8 is implemented.

Images