Multi-level protection method and system for wind turbine generator based on autonomous control

Abstract

The invention relates to the technical field of wind turbine generator control, and discloses a wind turbine generator multi-level protection method and system based on autonomous control, and the method comprises the steps: constructing a three-layer control architecture in a wind power plant, and respectively setting control nodes; a multi-layer redundant network is constructed through autonomous control nodes in the three-layer control architecture; real-time data are collected through a sensor, and the control nodes of all the control layers are monitored in real time. By constructing a three-layer control architecture including a main control layer, a regional control layer and a local control layer and combining a multi-layer redundant network of autonomous control nodes, efficient protection of the wind turbine generator is realized. When a certain control node fails, the system can quickly take over tasks, and continuous operation of the wind power plant is ensured. Real-time data are collected through the sensors, and the system can carry out fault prediction, anomaly detection and node failure judgment, so that the timeliness and accuracy of fault processing are improved, and the safety, stability and operation efficiency of the wind power plant are remarkably improved.

Description

Technical Field

[0001] The present invention relates to the technical field of wind turbine control, and in particular to a multi-level protection method and system for wind turbines based on autonomous control. Background Art

[0002] As the global demand for renewable energy continues to grow, wind power technology has developed rapidly as an important source of green energy. In modern wind farms, wind turbine control technology has evolved from the initial simple unit control to today's multi-level distributed control. Early wind turbines were usually managed by a single centralized control system, which was responsible for the operation monitoring, fault detection and power generation scheduling of the entire wind farm. However, with the expansion of wind farm scale and the complexity of technology, the single control architecture has gradually exposed its limitations, especially when dealing with the operation of multiple wind turbines in large-scale wind farms. The load and response capabilities of the centralized control system are limited, making it difficult to ensure efficient and reliable operation. In order to solve this problem, multi-level control architectures and distributed autonomous control systems have emerged in recent years, which can manage the operation of wind farms more flexibly. However, the existing technology still has certain deficiencies in redundant control, fault prediction and system stability, and it is difficult to fully meet the safety and efficiency requirements of future large-scale wind farms.

[0003] There are many aspects of existing wind turbine control technology that need to be improved. First, traditional redundant control mechanisms mostly rely on backup modules, which take over when a control unit fails. However, when there are many control levels and complex networks, the response speed and accuracy of redundant takeover are difficult to guarantee. In addition, current fault detection is mostly based on simple rules and fixed thresholds, which makes it difficult to accurately predict potential problems in wind turbine operation, resulting in the equipment failing to take preventive measures in time before failure. At the same time, the existing control architecture has limited control and coordination capabilities at the regional and local levels, and lacks effective mechanisms for real-time monitoring and fault self-repair. These problems greatly increase the maintenance cost of wind farms, and are prone to local or even global failure of the system in extreme climate or network outages.

[0004] Existing wind turbine control technologies generally face the problems of low redundant control efficiency, poor fault prediction accuracy and insufficient node failure management. Our invention proposes a multi-level protection method and system for wind turbines based on autonomous control, aiming to achieve real-time monitoring of control nodes, fault prediction and node failure judgment through a three-layer control architecture and a multi-level redundant network, combined with real-time data collected by sensors, thereby effectively solving the deficiencies in the existing technology. This invention belongs to the field of autonomous control of wind turbines, and is particularly suitable for the safe and efficient management of large wind farms. Summary of the invention

[0005] In view of the above-mentioned problems, the present invention is proposed.

[0006] Therefore, the technical problems solved by the present invention are: the low efficiency of redundant control, poor fault prediction accuracy, insufficient node failure management, and insufficient safety and stability of wind turbines in complex environments in the prior art.

[0007] In order to solve the above technical problems, the present invention provides the following technical solutions: a multi-level protection method for wind turbines based on autonomous control, comprising:

[0008] Construct a three-layer control architecture in the wind farm and set up control nodes separately;

[0009] Build a multi-level redundant network through autonomous control nodes in the three-layer control architecture;

[0010] Real-time data is collected through sensors to monitor the control nodes of each control layer in real time.

[0011] As a preferred solution of the multi-level protection method for wind turbines based on autonomous control described in the present invention, wherein: the three-layer control architecture includes a main control layer, a regional control layer and a local control layer;

[0012] The main control layer includes two main control nodes, which are responsible for centralized monitoring, dispatching and grid interaction of the entire farm, monitoring the status of all wind turbines, and issuing global instructions, and are located in the central control system of the wind farm;

[0013] The regional control layer includes three regional control nodes, which serve as an intermediate layer between the main control layer and the local control layer, perform regional management on the wind turbine group in charge, realize coordinated control of wind turbines in the region, collect data from the local control layer, and perform preliminary processing and aggregation;

[0014] The local control layer includes a local control node for each fan, which directly controls the operation of the single fan and collects and monitors the key parameters of the single fan in real time through sensors.

[0015] As a preferred solution of the multi-level protection method for wind turbines based on autonomous control described in the present invention, wherein: the control nodes include a master control node, a regional control node and a local control node;

[0016] The multi-level redundant network includes main control layer redundancy, regional control layer redundancy and local control layer redundancy.

[0017] As a preferred solution of the multi-level protection method of wind turbines based on autonomous control described in the present invention, wherein: the main control layer redundancy includes two main control nodes divided into a main node and a backup node. When the main node fails, the backup node takes over the global control task;

[0018] The regional control layer redundancy includes that the three regional control nodes can back up each other. When a regional control node fails, other regional control nodes can take over the control tasks of the failed area.

[0019] The local control layer redundancy includes that each local control node independently controls the wind turbine and makes autonomous decisions when the regional control fails.

[0020] As a preferred solution of the multi-level protection method of wind turbines based on autonomous control described in the present invention, the sensors include vibration sensors, temperature sensors, rotation speed sensors, wind speed sensors, wind direction sensors, voltage sensors, current sensors, pitch angle sensors, hydraulic sensors, acceleration sensors, inclination sensors and humidity sensors;

[0021] The real-time data includes fan operation status data, environmental data, mechanical system data, and electrical system data.

[0022] As a preferred solution of the multi-level protection method for wind turbines based on autonomous control described in the present invention, wherein: real-time monitoring of control nodes of each control layer includes using real-time data to monitor operating parameters, detect anomalies, predict faults, and determine node failures through a global monitoring module;

[0023] The operating parameter monitoring includes real-time monitoring of the fan operating data and comparing it with the set threshold value to ensure that the fan operates within a safe range;

[0024] The abnormality detection includes analyzing the changes in the historical records of sensor data and the real-time data to determine the abnormal situation.

[0025] As a preferred solution of the multi-level protection method for wind turbines based on autonomous control described in the present invention, the fault prediction includes dividing the historical data into a training set, a validation set and a test set; using an autoencoder model to train on the training set, perform tuning on the validation set, and perform a final evaluation on the test set;

[0026] The node failure includes that each self-control node periodically sends a heartbeat signal to the main control system and adjacent self-control nodes. When the main control system and adjacent self-control nodes do not receive the heartbeat signal of a node within a specified time, it is determined that the node has failed.

[0027] A multi-level protection system for wind turbines based on autonomous control, wherein:

[0028] Three-layer control architecture module, including main control layer, regional control layer and local control layer;

[0029] Redundant control module, which builds a multi-level redundancy mechanism at the main control layer, regional control layer and local control layer to achieve task switching, fault detection and self-repair, ensuring that the wind turbine system can still operate stably when the control node fails;

[0030] The fault prediction module uses an autoencoder model to predict faults. The model is trained on the training set, fine-tuned on the validation set, and evaluated on the test set.

[0031] The node failure judgment module detects whether a node is failed through the heartbeat signal mechanism. If a node does not send a heartbeat signal, the system automatically determines that the node is failed and starts redundant switching;

[0032] The global monitoring module is responsible for real-time monitoring of the overall operating parameters of the wind farm and resource scheduling to ensure that the wind turbines operate efficiently within a safe range.

[0033] A computer device comprises: a memory and a processor; the memory stores a computer program, wherein the processor implements the steps of any one of the methods of the present invention when executing the computer program.

[0034] A computer-readable storage medium stores a computer program, wherein the computer program, when executed by a processor, implements the steps of any one of the methods of the present invention.

[0035] Beneficial effects of the present invention: The multi-level protection method for wind turbines based on autonomous control provided by the present invention realizes efficient protection of wind turbines by constructing a three-layer control architecture, including a main control layer, a regional control layer, and a local control layer, combined with a multi-level redundant network of autonomous control nodes. When a control node fails, the system can quickly take over the task to ensure the continuous operation of the wind farm. In addition, by collecting real-time data through sensors, the system can perform fault prediction, anomaly detection, and node failure judgment, thereby improving the timeliness and accuracy of fault handling and significantly improving the safety, stability, and operating efficiency of the wind farm. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0037] Figure 1 An overall flow chart of a multi-level protection method for wind turbines based on autonomous control provided in the first embodiment of the present invention. DETAILED DESCRIPTION

[0038] In order to make the above-mentioned purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the drawings of the specification. Obviously, the described embodiments are part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary persons in the art without creative work should fall within the scope of protection of the present invention.

[0039] Example 1, reference Figure 1 , is an embodiment of the present invention, and provides a multi-level protection method for wind turbines based on autonomous control, comprising:

[0040] S1: Build a three-layer control architecture in the wind farm and set control nodes separately.

[0041] The three-layer control architecture includes a main control layer, a regional control layer and a local control layer.

[0042] The main control layer includes two main control nodes, which are responsible for centralized monitoring, dispatching and grid interaction of the entire field, monitoring the status of all wind turbines, and issuing global instructions. They are located in the central control system of the wind farm.

[0043] The regional control layer includes three regional control nodes, which serve as an intermediate layer between the main control layer and the local control layer, perform regional management of the wind turbine group they are responsible for, realize coordinated control of wind turbines in the region, collect data from the local control layer, and perform preliminary processing and aggregation.

[0044] The local control layer includes a local control node for each fan, which directly controls the operation of the single fan and collects and monitors the key parameters of the single fan in real time through sensors.

[0045] Furthermore, by building a three-layer architecture of the main control layer, regional control layer, and local control layer, the responsibilities of each layer are clarified to ensure hierarchical control of the wind farm. This hierarchical structure helps to improve the overall flexibility and response capabilities of the system and reduce the impact of a single failure on the entire wind farm. At the same time, by setting up multiple control nodes, it lays the foundation for the subsequent redundant control mechanism and enhances the stability and scalability of the system.

[0046] S2: Build a multi-level redundant network through autonomous control nodes in the three-layer control architecture.

[0047] The control nodes include a master control node, a regional control node and a local control node.

[0048] The multi-level redundant network includes main control layer redundancy, regional control layer redundancy and local control layer redundancy.

[0049] The main control layer redundancy includes two main control nodes, namely a main node and a backup node. When the main node fails, the backup node takes over the global control task;

[0050] The regional control layer redundancy includes that the three regional control nodes can back up each other. When a regional control node fails, other regional control nodes can take over the control tasks of the failed area.

[0051] The local control layer redundancy includes that each local control node independently controls the wind turbine and makes autonomous decisions when the regional control fails.

[0052] Furthermore, based on the three-layer control architecture, a redundant network is built through autonomous control nodes to achieve mutual backup and redundant control of each level of the system. When any node in the main control layer, regional control layer or local control layer fails, the redundant network can quickly take over the task to ensure the continuous operation of the wind turbine group or a single wind turbine. This design significantly improves the system's resilience in the event of a fault, and enhances the reliability and continuous operation of the wind farm.

[0053] S3: Collect real-time data through sensors and monitor the control nodes of each control layer in real time.

[0054] The sensors include a vibration sensor, a temperature sensor, a rotation speed sensor, a wind speed sensor, a wind direction sensor, a voltage sensor, a current sensor, a pitch angle sensor, a hydraulic sensor, an acceleration sensor, an inclination sensor and a humidity sensor.

[0055] The real-time data includes fan operation status data, environmental data, mechanical system data, and electrical system data.

[0056] Real-time monitoring of control nodes at each control layer includes using real-time data to monitor operating parameters, detect anomalies, predict faults, and determine node failures through a global monitoring module.

[0057] The operating parameter monitoring includes real-time monitoring of the fan operating data and comparing it with a set threshold value to ensure that the fan operates within a safe range.

[0058] The abnormality detection includes analyzing the changes in the historical records of sensor data and the real-time data to determine the abnormal situation.

[0059] The fault prediction includes dividing the historical data into a training set, a validation set and a test set;

[0060] The autoencoder model is trained on the training set, fine-tuned on the validation set, and finally evaluated on the test set. The formula is expressed as:

[0061]

[0062] Where x represents input data (such as sensor data), represents the data reconstructed by the autoencoder, represents the difference between the input data and the reconstructed data, δ represents the preset threshold, which represents the boundary between normal and faulty data, y represents the output of fault prediction, 1 represents a fault, and 0 represents no fault.

[0063] The node failure includes that each self-control node periodically sends a heartbeat signal to the main control system and adjacent self-control nodes. When the main control system and adjacent self-control nodes do not receive the heartbeat signal of a node within a specified time, it is determined that the node has failed.

[0064] Furthermore, the real-time operation data of the fan (such as speed, temperature, vibration, etc.) is collected through the sensor network, and combined with the global monitoring module to monitor the operation parameters, predict faults and determine node failures. The collection and monitoring of real-time data can not only ensure that the fan operates within a safe range, but also warn of potential faults in advance, greatly improving the safety of the system and the efficiency of fault response.

[0065] Embodiment 2 is an embodiment of the present invention, which provides a multi-level protection system for wind turbines based on autonomous control, including:

[0066] The three-layer control architecture module includes the main control layer, the regional control layer and the local control layer.

[0067] The redundant control module builds a multi-level redundancy mechanism at the main control layer, regional control layer and local control layer to achieve task switching, fault detection and self-repair, ensuring that the wind turbine system can still operate stably when the control node fails.

[0068] The fault prediction module performs fault prediction through the autoencoder model. The model is trained on the training set, fine-tuned on the validation set, and evaluated on the test set.

[0069] The node failure judgment module detects whether a node is failed through the heartbeat signal mechanism. If a node does not send a heartbeat signal, the system automatically determines that the node is failed and starts redundant switching.

[0070] The global monitoring module is responsible for real-time monitoring of the overall operating parameters of the wind farm and resource scheduling to ensure that the wind turbines operate efficiently within a safe range.

[0071] Embodiment 3, an embodiment of the present invention, is different from the first two embodiments in that:

[0072] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc., which can store program codes.

[0073] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in conjunction with such instruction execution systems, devices or apparatuses. For the purposes of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in conjunction with such instruction execution systems, devices or apparatuses.

[0074] More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection with one or more wires (electronic device), a portable computer disk case (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be a paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, deciphering or, if necessary, processing in another suitable manner, and then stored in a computer memory.

[0075] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, a plurality of steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0076] Embodiment 4 adopts a simulated wind farm environment to construct a three-layer control architecture, including a main control layer, a regional control layer and a local control layer, and sets 2 main control nodes, 3 regional control nodes and 15 local control nodes respectively. The local control node is directly connected to the wind turbine, and the real-time operating status data of the wind turbine is collected through vibration sensors, temperature sensors, wind speed sensors, etc. The system monitors the operating status of each control node in real time through a global monitoring module, and performs fault prediction and node failure judgment in combination with the autoencoder model. In order to evaluate the effectiveness of the inventive method, the data is divided into a training set, a validation set and a test set, and the autoencoder model is trained, tuned and tested to analyze the accuracy of its fault prediction and node failure detection.

[0077] At the beginning of the test, all wind turbines operated normally, and each node of the three-layer control architecture worked normally according to the predetermined function. The global monitoring module began to collect the operating data of the wind turbine in real time and tracked the status of each node. During the test, node failures were randomly introduced in the local control layer, regional control layer and main control layer to simulate common fault scenarios, such as local node wind turbine failure, regional controller disconnection and main control node overload failure. In each simulated scenario, the redundant control mechanism was started, and the backup node or adjacent node took over the control task to ensure the continuous operation of the wind farm as a whole. After receiving the collected real-time data, the autoencoder model automatically analyzes the reconstruction error of each node. When the reconstruction error exceeds the threshold, the system outputs a fault prediction signal and starts the redundant switching mechanism in the case of node failure to ensure the smooth operation of the system. By comparing the existing traditional centralized control system, the response time, node recovery time, redundant switching efficiency and fault prediction accuracy of the wind turbine in the simulated fault scenario are recorded to analyze the performance of the invented system. The experimental data are shown in Table 1.

[0078] Table 1 Experimental data table

[0079]

[0080]

[0081] It can be clearly seen from the table data that the multi-level protection method for wind turbines based on autonomous control of the present invention is superior to the traditional centralized control system in various indicators, especially in terms of response time, redundancy switching success rate and fault prediction accuracy. In the test of multiple wind turbines, the response time of the system for wind turbines A and E in the local control layer was 0.8 seconds and 0.7 seconds respectively, which is significantly better than the 2.5 seconds of the traditional system. This shows that the multi-level control architecture of the present invention can respond quickly to node failures and realize seamless switching of tasks through a redundant mechanism. The redundancy switching success rate of all wind turbines is 100%, which proves the excellent performance of the redundant control module of the present invention in ensuring system continuity.

[0082] In terms of fault prediction, the system based on the autoencoder model can efficiently analyze the operating data of each fan and accurately predict possible faults. In the experiment, the fault prediction accuracy of fans A, B, C and E exceeded 95%, while the fault prediction accuracy of the traditional system was only 85%. This comparison result proves the advantage of the autoencoder model, which can identify potential faults in advance and issue early warnings by learning from historical data and analyzing real-time data, avoiding false alarms or missed alarms caused by fixed threshold judgments in traditional systems.

[0083] The comparison of node failure rates shows that the failure rate of wind turbines in the present invention is relatively low, and even if a node fails, the system's redundancy mechanism can take over the task in time, ensuring that the overall system stability is above 98%, which is significantly higher than the 90% of the traditional system. This difference reflects the effectiveness of the multi-level redundant network in preventing node failures.

[0084] Through these data analysis, it can be concluded that the multi-level protection method of the present invention significantly improves the safety and operating efficiency of the wind turbine system, especially in fault prediction and redundant switching. It shows great innovation and novelty. Compared with the existing technology, the present invention not only solves the problems of slow response and low prediction accuracy of traditional control systems, but also realizes efficient and reliable wind turbine protection through the combination of redundant control mechanism and autoencoder algorithm, and has broad application prospects.

[0085] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A multi-level protection method for wind turbines based on autonomous control, characterized in that: include: Construct a three-layer control architecture in the wind farm and set up control nodes separately; Build a multi-level redundant network through autonomous control nodes in the three-layer control architecture; Real-time data is collected through sensors to monitor the control nodes of each control layer in real time.

2. The multi-level protection method for wind turbines based on autonomous control as claimed in claim 1, characterized in that: The three-layer control architecture includes a main control layer, a regional control layer and a local control layer; The main control layer includes two main control nodes, which are responsible for centralized monitoring, dispatching and grid interaction of the entire farm, monitoring the status of all wind turbines, and issuing global instructions, and are located in the central control system of the wind farm; The regional control layer includes three regional control nodes, which serve as an intermediate layer between the main control layer and the local control layer, perform regional management on the wind turbine group in charge, realize coordinated control of wind turbines in the region, collect data from the local control layer, and perform preliminary processing and aggregation; The local control layer includes a local control node for each fan, which directly controls the operation of the single fan and collects and monitors the key parameters of the single fan in real time through sensors.

3. The multi-level protection method for wind turbines based on autonomous control as claimed in claim 1, characterized in that: The control nodes include a master control node, a regional control node and a local control node; The multi-level redundant network includes main control layer redundancy, regional control layer redundancy and local control layer redundancy.

4. The multi-level protection method for wind turbines based on autonomous control as claimed in claim 3, characterized in that: The main control layer redundancy includes two main control nodes, namely a main node and a backup node. When the main node fails, the backup node takes over the global control task; The regional control layer redundancy includes that the three regional control nodes can back up each other. When a regional control node fails, other regional control nodes can take over the control tasks of the failed area. The local control layer redundancy includes that each local control node independently controls the wind turbine and makes autonomous decisions when the regional control fails.

5. The multi-level protection method for wind turbines based on autonomous control according to claim 1, characterized in that: The sensors include vibration sensors, temperature sensors, rotation speed sensors, wind speed sensors, wind direction sensors, voltage sensors, current sensors, pitch angle sensors, hydraulic sensors, acceleration sensors, inclination sensors and humidity sensors; The real-time data includes fan operation status data, environmental data, mechanical system data, and electrical system data.

6. The multi-level protection method for wind turbines based on autonomous control as claimed in claim 5, characterized in that: Real-time monitoring of control nodes at each control layer includes using real-time data to monitor operating parameters, detect anomalies, predict faults, and determine node failures through a global monitoring module; The operating parameter monitoring includes real-time monitoring of the fan operating data and comparing it with the set threshold value to ensure that the fan operates within a safe range; The abnormality detection includes analyzing the changes in the historical records of sensor data and the real-time data to determine the abnormal situation.

7. The multi-level protection method for wind turbines based on autonomous control as claimed in claim 6, characterized in that: The fault prediction includes dividing the historical data into a training set, a validation set and a test set; Use the autoencoder model to train on the training set, fine-tune on the validation set, and perform final evaluation on the test set; The node failure includes that each self-control node periodically sends a heartbeat signal to the main control system and adjacent self-control nodes. When the main control system and adjacent self-control nodes do not receive the heartbeat signal of a node within a specified time, it is determined that the node has failed.

8. A multi-level protection system for wind turbines based on autonomous control using the method according to any one of claims 1 to 7, characterized in that: Three-layer control architecture module, including main control layer, regional control layer and local control layer; Redundant control module, which builds a multi-level redundancy mechanism at the main control layer, regional control layer and local control layer to achieve task switching, fault detection and self-repair, ensuring that the wind turbine system can still operate stably when the control node fails; The fault prediction module uses an autoencoder model to predict faults. The model is trained on the training set, fine-tuned on the validation set, and evaluated on the test set. The node failure judgment module detects whether a node is failed through the heartbeat signal mechanism. If a node does not send a heartbeat signal, the system automatically determines that the node is failed and starts redundant switching; The global monitoring module is responsible for real-time monitoring of the overall operating parameters of the wind farm and resource scheduling to ensure that the wind turbines operate efficiently within a safe range.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the multi-level protection method for wind turbines based on autonomous control as described in any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the multi-level protection method for wind turbines based on autonomous control as described in any one of claims 1 to 7 are implemented.

Images