Edge field collaborative control system and method of wind farm, wind turbine generator, medium and program product

Abstract

The application discloses a wind farm edge field cooperative control system and method, a wind turbine generator unit, a medium and a program product, and belongs to the field of wind power generation. The system comprises: a unit side control device and a field level control system. The unit side control device generates a single-machine candidate control instruction according to collected state monitoring data and working condition data of the wind turbine generator unit, and transmits the state monitoring data and the working condition data to the field level control system. The field level control system generates a field level candidate control instruction according to the state monitoring data and the working condition data. The unit side control device does not unconditionally execute the field level control instruction, but reasonably analyzes the single-machine candidate control instruction and the field level candidate control instruction, determines a target control instruction and executes the same. The embodiments of the application are used for wind farm control, can avoid that an abnormal control instruction generated by the field level control system due to an anomaly is executed by the wind turbine generator unit, avoid that an over-limit action is executed by an executing mechanism, and improve the safety of wind farm control.

Description

Technical Field

[0001] This application belongs to the field of wind power generation, and in particular relates to a side-field coordinated control system, method, wind turbine, medium and program product for a wind farm. Background Technology

[0002] A wind farm may include multiple wind turbines, wind turbine control devices, a Supervisory Control and Data Acquisition (SCADA) system, a condition monitoring system, a substation centralized control system, and an operation and maintenance platform. Through the collaboration of these structures, individual turbine control, condition monitoring, farm-wide monitoring, and wake optimization control can be achieved within the wind farm. Wind farms can employ a farm-wide centralized control approach. The farm-wide control system generates control instructions based on the acquired data from each wind turbine and sends these instructions to the control devices of each turbine. The control devices of each turbine respond to the control execution, driving the turbine actuators to perform corresponding actions. However, the farm-wide control system may generate abnormal control instructions due to software defects, hardware failures, model drift, misconfiguration, bad data, etc., and send these instructions to the wind turbine control devices. The wind turbine control devices responding to these abnormal instructions may cause the actuators to perform actions beyond their limits, resulting in abnormal wind turbine operation or even damage to the turbines, thus reducing the safety of wind farm control. Summary of the Invention

[0003] This application provides a wind farm edge-field coordinated control system, method, wind turbine, medium, and program product that can prevent actuators from performing out-of-bounds actions and improve the safety of wind farm control.

[0004] In a first aspect, embodiments of this application provide a side-field coordinated control system for a wind farm, comprising: a turbine-side control device, corresponding to a wind turbine; and a farm-level control system, communicatively connected to the turbine-side control device; wherein the turbine-side control device generates a single-unit candidate control instruction based on the collected status monitoring data and operating condition data of the wind turbine, and transmits the status monitoring data and operating condition data to the farm-level control system; the farm-level control system generates a farm-level candidate control instruction based on the status monitoring data and operating condition data; and the turbine-side control device determines a target control instruction and executes the target control instruction based on the single-unit candidate control instruction and the farm-level candidate control instruction.

[0005] In some possible embodiments, the unit-side control device includes: a wind turbine control device; and a unit edge control device communicatively connected to the wind turbine control device; wherein the wind turbine control device provides operating condition data to the unit edge control device; the unit edge control device generates single-unit candidate control instructions based on the status monitoring data and operating condition data, and transmits the status monitoring data and operating condition data to the farm-level control system; the unit edge control device determines a target control instruction based on the single-unit candidate control instruction and the farm-level candidate control instruction, and assigns it to the wind turbine control device so that the wind turbine control device executes the target control instruction.

[0006] In some possible embodiments, the system further includes: a data isolation communication device for connecting the unit edge control device and the field-level control system, the data isolation communication device including a first data channel and a second data channel, the first data channel for transmitting data from the unit edge control device to the field-level control system, and the second data channel for transmitting data from the field-level control system to the unit edge control device.

[0007] In some possible embodiments, the unit edge control device includes: a first control module, configured to generate a single-unit candidate control instruction based on status monitoring data, operating condition data, and single-unit optimized control objectives, and to obtain a target control instruction based on the single-unit candidate control instruction and the field-level candidate control instruction; a second control module, wherein an independent control channel is provided between the second control module and the wind turbine control device, or an independent control channel is provided between the second control module and the actuator; and, upon triggering a first predetermined condition, the second control module transmits a control instruction for the first predetermined condition through the independent control channel, wherein the first predetermined condition is used to determine that an emergency safety event has occurred.

[0008] In some possible embodiments, the unit edge control device is used to acquire safety control auxiliary data, and based on the single-unit candidate control indication, the safety control auxiliary data, and the farm-level candidate control indication, to determine the allocation of target control indications to the wind turbine control device, wherein the safety control auxiliary data is used to help determine the relative importance of the single-unit candidate control indication and the farm-level candidate control indication.

[0009] In some possible embodiments, safety control auxiliary data includes influencing factors that can affect the importance of individual unit candidate control indications and farm-level candidate control indications in wind turbine control.

[0010] In some possible embodiments, security control auxiliary data includes at least one of security status, communication status, single-machine preset operating boundary, single-machine security envelope, indication verification result, channel integrity monitoring result, and trustworthiness data; Safety status is used to characterize whether a safety event has been triggered; communication status is used to characterize the communication status between the unit edge control device and the field-level control system, and / or the communication status between the second control module in the unit edge control device and the wind turbine control device; the single-unit preset operating boundary is the boundary of the operating data that the wind turbine needs to follow; the single-unit safety envelope is the dynamic safety boundary of the wind turbine's operating data obtained based on status monitoring data, operating condition data, and the lightweight safety simulation sub-model in the unit edge control device; the indication verification result is the verification result of the field-level candidate control indication; the channel integrity monitoring result is used to characterize whether the data channels in the data isolation communication device and the data channels in the independent control channels diverge, and the data channels include the first data channel and the second data channel; the credibility data is used to characterize the credibility of any one or more of the data acquisition channels, operating condition data, sub-models, and field-level candidate control indications of the data acquisition device, and the sub-models include the lightweight safety simulation sub-model and multiple physical sub-models related to the wind farm and wind turbines set in the field-level control system.

[0011] In some possible embodiments, in the event of a unit decision triggering event, the first data channel changes the data transmission parameters. The unit decision triggering event indicates that the change in the state of the wind turbine exceeds the normal range.

[0012] In some possible embodiments, the unit decision-triggered event includes one or more of the following: a safety event, a change in operating condition data exceeding a first change threshold, a change in condition monitoring data exceeding a second change threshold, and a change in the reliability represented by reliability data exceeding a third change threshold.

[0013] In some possible embodiments, the unit edge control device is used to: generate a target control instruction based on the field-level candidate control instruction when the safety control auxiliary data and the field-level candidate control instruction do not meet the degradation control conditions; and reject the field-level candidate control instruction and determine the target control instruction based on the met degradation control conditions when the safety control auxiliary data or the field-level candidate control instruction meets the degradation control conditions.

[0014] In some possible embodiments, the unit edge control device is provided with multiple degradation control conditions and their corresponding control instructions. The degradation control conditions are related to the degree of deterioration of the wind turbine's state, and different degradation control conditions have corresponding recovery control conditions. The unit edge control device is also used to: after executing the target control instruction corresponding to the met degradation control conditions, if the status monitoring data and operating condition data continue to meet the current recovery control conditions within a preset time period, execute the control instruction corresponding to the degradation control conditions with a less severe degree of state deterioration compared to the met degradation control conditions.

[0015] In some possible embodiments, the unit edge control device is specifically used for one or more of the following scenarios: if the safety status representation in the safety control auxiliary data triggers a safety event, then the field-level candidate control indication is rejected; if the channel integrity monitoring result in the safety control auxiliary data represents overlapping data divergence between the data channel in the data isolation communication device and the independent safety control channel, the communication status represents communication interruption, or the indication verification result represents failure to verify the field-level candidate control indication, then the field-level candidate control indication is rejected; if any control quantity in the field-level candidate control indication exceeds the single-machine safety envelope in the safety control auxiliary data, then the target control indication is determined based on the projection of the field-level candidate indication onto the single-machine safety envelope; if the credibility of the credibility data representation in the safety control auxiliary data is lower than the credibility threshold, then the field-level candidate control indication is rejected.

[0016] In some possible embodiments, the wind farm-level control system is further configured to: upon receiving a safety event sent by the edge control device of any wind turbine in the wind farm, determine the affected turbine set based on the safety event and the correlation between wind turbines in the wind farm, the affected turbine set including the wind turbines affected by the wind turbine in which the safety event occurred; generate linkage candidate control indications based on the severity of the safety event, and assign linkage candidate control indications to the edge control devices of the wind turbines in the affected turbine set; collect status monitoring data and operating condition data of the wind turbines in the affected turbine set; and, if the collected status monitoring data and operating condition data do not meet the predetermined recovery judgment conditions, assign linkage candidate control indications with higher severity to the edge control devices of the wind turbines in the affected turbine set.

[0017] In some possible embodiments, the unit edge control device is also used to: determine the target linkage control instruction and assign it to the unit control device based on the single-unit candidate control instruction, safety control auxiliary data and linkage candidate control instruction.

[0018] In some possible embodiments, the wind turbines in the wind farm are divided into multiple wind turbine clusters, and the farm-level control system includes: multiple cluster servers, each corresponding to a wind turbine cluster, which are used to generate cluster candidate control instructions based on status monitoring data and operating condition data; and a farm-level server, which is communicatively connected to the cluster servers and is used to generate farm-level candidate control instructions based on the cluster candidate control instructions generated by each cluster server.

[0019] Secondly, embodiments of this application provide a side-field coordinated control method for a wind farm, applied to the side-field coordinated control system of the wind farm in the first aspect. The method includes: a turbine-side control device acquiring status monitoring data and operating condition data of the wind turbine, and transmitting the status monitoring data and operating condition data to the farm-level control system; the farm-level control system generating farm-level candidate control instructions based on the status monitoring data and operating condition data, and issuing them to the turbine-side control device; the turbine-side control device determining a target control instruction and executing the target control instruction based on the single-unit candidate control instruction and the farm-level candidate control instruction, wherein the single-unit candidate control instruction is generated by the turbine-side control device based on the status monitoring data and operating condition data.

[0020] Thirdly, embodiments of this application provide a wind turbine generator, including: an electronic device, the electronic device including a processor and a memory storing computer program instructions; when the processor executes the computer program instructions, it implements the wind farm side-field coordinated control method of the second aspect.

[0021] Fourthly, embodiments of this application provide a computer-readable storage medium storing computer program instructions, which, when executed by a processor, implement the wind farm side-field coordinated control method of the second aspect.

[0022] Fifthly, embodiments of this application provide a computer program product, including a computer program that, when executed by a processor, implements the wind farm side-field coordinated control method of the second aspect.

[0023] This application provides a wind farm side-field coordinated control system, method, wind turbine, medium, and program product. The turbine-side control device can acquire the status monitoring data and operating condition data of the wind turbine, and generate single-unit candidate control instructions based on the status monitoring data and operating condition data. The farm-level control system acquires the status monitoring data and operating condition data from the turbine-side control device, generates farm-level candidate control instructions based on the status monitoring data and operating condition data, and sends the farm-level candidate control instructions to the turbine-side control device. The turbine-side control device analyzes the rationality of the candidate instructions based on the single-unit candidate control instructions and the farm-level candidate control instructions, thereby determining a reasonable target control instruction and executing the target control instruction. The turbine-side control device does not unconditionally execute the farm-level control instructions issued by the farm-level control system. Instead, it analyzes the rationality of the farm-level candidate control instructions to determine the target control instruction to be executed by the wind turbine. This avoids abnormal control instructions from the farm-level control system due to software defects, hardware failures, model drift, misconfiguration, bad data, etc., which could be executed by the wind turbine. Consequently, it prevents the actuator from performing out-of-bounds actions, ensuring the normal operation of the wind turbine and improving the safety of wind farm control. Attached Figure Description

[0024] To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings used in the embodiments of this application will be briefly introduced below. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0025] Figure 1 A schematic diagram of the structure of a wind farm sidefield coordinated control system provided in an embodiment of this application; Figure 2 A schematic diagram of the structure of a wind farm sidefield coordinated control system provided in another embodiment of this application; Figure 3 A schematic diagram of the structure of a wind farm sidefield coordinated control system provided in another embodiment of this application; Figure 4 A schematic diagram of an example of a unit edge control device provided in an embodiment of this application; Figure 5 This is a schematic diagram illustrating an example of the interaction between the unit edge control device and the wind turbine control device provided in an embodiment of this application. Figure 6 A schematic diagram illustrating an example of the process for determining credibility data provided in an embodiment of this application; Figure 7 A schematic diagram illustrating an example of the gain for determining the model parameter update of a physical sub-model, as provided in an embodiment of this application; Figure 8 A flowchart illustrating an example of determining a target control instruction through degraded control conditions, provided in an embodiment of this application; Figure 9 A schematic diagram illustrating an example of a state machine provided in an embodiment of this application; Figure 10 A flowchart illustrating an example of field-level safety linkage provided in an embodiment of this application; Figure 11 A flowchart illustrating an example of wind farm safety linkage processing and data collection verification provided in this application embodiment; Figure 12 A flowchart of a wind farm sidefield coordinated control method provided in an embodiment of this application. Detailed Implementation

[0026] The features and exemplary embodiments of various aspects of this application will be described in detail below. To make the purpose, technical solution, and advantages of this application clearer, the application will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are only intended to explain this application and not to limit it. For those skilled in the art, this application can be implemented without some of the details in these specific details. The following description of the embodiments is merely to provide a better understanding of this application by illustrating examples of this application. It should be noted that the acquisition, storage, use, and processing of information and data in the embodiments of this application are all authorized by users or relevant organizations and comply with the relevant provisions of national laws and regulations. In the embodiments of this application, certain software, components, models, and other existing solutions in the industry may be mentioned. These should be considered as exemplary, and their purpose is only to illustrate the feasibility of implementing the technical solution of this application, but it does not mean that the applicant has or necessarily used such a solution.

[0027] A wind farm may include multiple wind turbines, their control devices, a SCADA system, a condition monitoring system, a substation centralized control system, and an operation and maintenance platform. Through the collaboration of these structures, individual turbine control, condition monitoring, farm-wide monitoring, and wake optimization control can be achieved. Wind farms can employ a farm-wide centralized control approach. The farm-wide control system generates control instructions based on acquired data from each wind turbine and sends these instructions to the control devices of each turbine. The control devices of each turbine respond to the control, driving the turbine's actuators to perform corresponding actions. However, the farm-wide control system may generate abnormal control instructions and send them to the turbine control devices due to software defects, hardware failures, model drift, misconfiguration, bad data, miscalculations, communication failures, or intrusion. The turbine control devices responding to these abnormal instructions may cause the actuators to perform actions beyond their limits, resulting in abnormal turbine operation or even damage, thus reducing the safety of wind farm control. Furthermore, in some cases, both the farm-level control system and the wind turbine control devices are deployed in the production control area. The farm-level control system within the production controller and the wind turbine control devices are only logically isolated, lacking strong constraints that prevent the wind turbine from crossing boundaries. If data diodes or unidirectional channels are installed between the farm-level control system and the wind turbine to isolate conduction, it will prevent the implementation of farm-level closed-loop control of the wind farm.

[0028] This application provides a wind farm side-field coordinated control system, method, wind turbine, medium, and program product. The turbine-side control device is communicatively connected to the farm-level control system. The farm-level control indication generated by the farm-level control system serves as a candidate indication, and the individual turbine control indication generated by the turbine-side control device also serves as a candidate indication. The turbine-side control device analyzes these indications to determine their rationality, and then determines the final target control indication to be executed based on the farm-level and individual turbine control indications. By analyzing the rationality of the farm-level control indications through the turbine-side control device, abnormal control indications generated by the farm-level control system due to software defects, hardware failures, model drift, misconfiguration, bad data, etc., are avoided from being executed by the wind turbine. A data isolation communication device can also be set up between the turbine-side control device and the farm-level control system to isolate them. Since the final target control indication is determined by the turbine-side control device, the farm-level closed-loop control can be disconnected at the network boundary between the turbine-side control device and the farm-level control system, and closed locally on the turbine-side control device, thereby simultaneously satisfying industrial control safety isolation and farm-level closed-loop control optimization.

[0029] The following describes the wind farm side-field coordinated control system, method, wind turbine, medium, and program products provided in this application.

[0030] This application provides a side-field coordinated control system for a wind farm. Figure 1 This is a schematic diagram of the structure of a wind farm side-field coordinated control system provided in an embodiment of this application, as shown below. Figure 1 As shown, the side-field coordinated control system of the wind farm may include a turbine-side control device 10 and a field-level control system 13.

[0031] The turbine-side control device 10 is configured correspondingly to the wind turbine and deployed on the wind turbine side. It can be used for turbine condition sensing, data evaluation, single-unit control optimization, control instruction determination, safety control, logic control execution, safety monitoring, and instruction execution. The turbine-side control device 10 can acquire one or more types of condition monitoring data from the wind turbine. In some examples, multiple sensors can be installed on the wind turbine to collect various types of condition monitoring data, and the turbine-side control device 10 can acquire the condition monitoring data from the sensors. For example, ... Figure 1As shown, the turbine-side control device 10 can communicate with a multi-source sensor group 16, which may include various sensors for collecting various status monitoring data. The status monitoring data can be used to monitor the status of the wind turbine. For example, the status monitoring data may include, but is not limited to, one or more of vibration data, temperature data, strain data, electrical data, meteorological data, and oil data. In some examples, the wind farm's side-field coordinated control system may also include the multi-source sensor group 16. The turbine-side control device 10 can acquire the wind turbine's operating condition data. The operating condition data can characterize the wind turbine's operating conditions. For example, the operating condition data may include, but is not limited to, one or more of turbine status data, power data, pitch data, and yaw data. The turbine status data can characterize the status of the wind turbine, which may include, but is not limited to, shutdown status, standby status, maintenance status, load reduction status, and normal operation status.

[0032] The wind farm-level control system 13 is configured correspondingly to the wind farm and deployed on the wind farm side. It is communicatively connected to the turbine-side control device 10 and can be used for data fusion of multiple wind turbines, wind farm-level control optimization, and wind farm-level safety decision-making. In some examples, the wind farm-level control system 13 may include, but is not limited to, a wind farm-level centralized control cabinet (FCC), a booster station server, a wind farm private cloud, and a wind farm-level control server cluster. The wind farm-level control system 13 may be deployed in the wind farm's booster station, central control room, or wind farm private cloud, etc., and is not limited here.

[0033] The unit-side control device 10 can generate single-unit candidate control instructions based on the collected various status monitoring data and operating condition data of the wind turbine, and transmit the status monitoring data and operating condition data to the field-level control system 13; the field-level control system 13 generates field-level candidate control instructions based on the status monitoring data and operating condition data; the unit-side control device 10 determines the target control instruction based on the single-unit candidate control instructions and the field-level candidate control instructions, and executes the target control instruction.

[0034] In this embodiment, the unit-side control device 10 can acquire the status monitoring data and operating condition data of the wind turbine, and generate single-unit candidate control instructions based on the status monitoring data and operating condition data; the farm-level control system 13 acquires the status monitoring data and operating condition data from the unit-side control device 10, generates farm-level candidate control instructions based on the status monitoring data and operating condition data, and sends the farm-level candidate control instructions to the unit-side control device 10; the unit-side control device 10 analyzes the rationality of the candidate instructions based on the single-unit candidate control instructions and the farm-level candidate control instructions, thereby determining a reasonable target control instruction and executing the target control instruction. The unit-side control device 10 does not unconditionally execute the field-level control instructions issued by the field-level control system 13. Instead, the unit-side control device 10 analyzes the rationality of the field-level candidate control instructions to determine the target control instructions to be executed by the wind turbine. This avoids the wind turbine from executing abnormal control instructions generated by the field-level control system 13 due to software defects, hardware failures, model drift, misconfiguration, bad data, or other anomalies. In this way, the actuator is prevented from performing out-of-bounds actions, ensuring the normal operation of the wind turbine and improving the safety of wind farm control.

[0035] In some embodiments, the unit-side control device 10 includes a wind turbine control device and a unit edge control device. Figure 2 A schematic diagram of the structure of a wind farm side-field coordinated control system provided in another embodiment of this application is shown below. Figure 2 As shown, the unit-side control device 10 includes a unit edge control device 11 and a wind turbine control device 12.

[0036] The turbine edge control device 11 is deployed on the wind turbine side and can be used for data evaluation, single-unit control optimization, control instruction determination, and safety control. In some examples, the turbine edge control device 11 may include, but is not limited to, one or more structures such as an edge box controller (EBC), a tower edge server, a nacelle edge controller, and a main control cabinet expansion module. The turbine edge control device 11 can acquire various status monitoring data from the wind turbine. In some examples, multiple sensors can be installed on the wind turbine to collect various status monitoring data, and the turbine edge control device 11 can acquire status monitoring data from the sensors. For example, such as... Figure 2 As shown, the unit edge control device 11 can communicate with the multi-source sensor group 16.

[0037] The wind turbine control device 12 is configured correspondingly to the wind turbine and deployed on the wind turbine side. It can be used for turbine operating condition sensing, logic control execution, safety monitoring, and issuing instructions to actuators. In some examples, the wind turbine control device 12 may include, but is not limited to, the programmable logic controller (PLC) of the wind turbine. The wind turbine control device 12 can acquire the operating condition data of the wind turbine. The wind turbine control device 12 is communicatively connected to the turbine edge control device 11, and can provide the sensed operating condition data to the turbine edge control device 11. The wind turbine control device 12 can also be connected to the actuator 15 of the wind turbine to drive the actuator 15 corresponding to the control instruction upon receiving a control instruction, so that the actuator 15 can be activated. In some examples, the wind farm's edge-field coordinated control system may also include the actuator 15, which has no direct connection to the farm-level control system 13. The actuator 15 is used to perform corresponding actions under the control of the wind turbine control device 12. For example, the actuator 15 may include, but is not limited to, one or more of the following: pitch system, converter, yaw system, shutdown system, load reduction system, etc.

[0038] The field-level control system 13 can communicate with the unit edge control device 11. The field-level control system 13 can obtain status monitoring data and operating condition data from the unit edge control device 11. The field-level control system 13 can also issue field-level control instructions to the unit edge control device 11 so that the unit edge control device 11 can determine the final target control instruction to be executed.

[0039] The wind turbine control unit 12 provides the wind turbine's operating condition data to the unit edge control unit 11. The unit edge control unit 11 generates individual unit candidate control instructions based on the collected various status monitoring data and operating condition data of the wind turbine, and transmits the status monitoring data and operating condition data to the farm-level control system 13. The farm-level control system 13 generates farm-level candidate control instructions based on the status monitoring data and operating condition data. Based on the individual unit candidate control instructions and the farm-level candidate control instructions, the unit edge control unit 11 determines the target control instruction and assigns it to the wind turbine control unit 12.

[0040] The unit edge control device 11 has its own single-unit control optimization logic. Based on status monitoring data, operating condition data, and its own single-unit control optimization logic, the unit edge control device can generate single-unit candidate control instructions that can achieve the optimization objectives in the single-unit control optimization logic. The optimization objectives in the single-unit control optimization logic can be configured according to specific scenarios and requirements, and are not limited here. For example, the optimization objectives in the single-unit control optimization logic may include achieving the required power generation of a single wind turbine, achieving a balance between the power generation and safety of a single wind turbine, or achieving the required safety of a single wind turbine, meeting grid dispatch constraints, and meeting load constraints, etc., and are not limited here. The single-unit candidate control instructions can indicate control parameters, which may include, but are not limited to, one or more of the following: pitch angle, generator torque, yaw angle, power limit value, load reduction ratio, etc. The single-unit candidate control instructions are candidate instructions, and the unit edge control device 11 needs to analyze and determine whether to adopt them. The unit edge control device 11 can also upload status monitoring data and operating condition data to the field-level control system 13.

[0041] The wind farm-level control system 13 has its own wind farm-level control optimization logic. Based on the status monitoring data and operating data of the wind turbines in the wind farm, and its own wind farm-level control optimization logic, it can generate wind farm-level candidate control instructions that can achieve the optimization objectives in the wind farm-level control optimization logic. The optimization objectives in the wind farm-level control optimization logic may include wind farm power generation reaching demand standards, wind farm power generation and wind farm safety reaching a balance or wind farm safety reaching demand standards, meeting grid dispatch constraints, and meeting load constraints, etc., and are not limited here. Achieving a balance between wind farm power generation and wind farm safety, or meeting wind farm safety reaching demand standards, means, under the condition of meeting power generation and grid constraints, redistributing the load or offloading among the wind turbines in the wind farm to balance the remaining lifespan of the wind turbines in the wind farm. For example, the wind farm-level control system 13 can generate wind farm-level candidate control instructions based on wind turbine status data, available power, load status, communication status, wind direction, wind speed, wake topology, grid dispatch constraints, and the version of the wind farm-level control optimization logic. The field-level candidate control indication can indicate control parameters, which may include, but are not limited to, one or more of the following: wind turbine yaw angle, power distribution value, load reduction ratio, wind turbine shutdown, wind turbine startup, shutdown time, shutdown batch, startup sequence, inspection strategy for a specific unit, power limitation for a specific unit, and lockout strategy for a specific unit. The field-level candidate control indication is a candidate indication, and the unit edge control device 11 needs to analyze it to determine whether to adopt it. The field-level control system 13 sends the field-level candidate control indication to the corresponding wind turbine unit's unit edge control device 11. In some examples, such as... Figure 2As shown, the field-level control system 13 may include a data access module 131, a mirror modeling module 132, and a field-level optimization module 133. The data access module 131 can be used to receive status monitoring data transmitted from the unit edge control device 11 and local operating condition data. The mirror modeling module 132 can be used to obtain the mirror state of the wind farm based on multiple physical sub-models. The physical sub-models are sub-models of the physical structure of the wind turbine units in the wind farm and sub-models of the physical structure of the wind farm. For example, the physical sub-models may include, but are not limited to, one or more of the following: aerodynamic sub-model, transmission chain sub-model, electromagnetic sub-model, tower structure sub-model, wind farm unit distribution model, wake model, etc. The field-level optimization module 133 can be used to generate field-level candidate control indications based on the mirror state of the wind farm.

[0042] The unit edge control device 11 has its own rationality analysis logic. Based on the single-unit candidate control indication, the farm-level candidate control indication, and the rationality analysis logic, it analyzes the rationality of the farm-level candidate control indication, thereby determining the target control indication to be assigned to the wind turbine control device 12. The target control indication is the control indication assigned to the wind turbine control device 12 for execution. The target control indication may include single-unit candidate control indications, farm-level candidate control indications, control indications obtained by fusing single-unit and farm-level candidate control indications, or other control indications, which are not limited here. Through the rationality analysis of the unit edge control device 11, abnormal control indications can be avoided from being issued as target control indications to the wind turbine control device 12 for execution, thereby preventing the actuator 15 from performing out-of-bounds actions and ensuring the normal operation of the wind turbine.

[0043] In this embodiment, the unit edge control device 11 can acquire the status monitoring data and operating condition data of the wind turbine, and generate a single-unit candidate control instruction based on the status monitoring data and operating condition data; the farm-level control system 13 acquires the status monitoring data and operating condition data from the unit edge control device 11, generates a farm-level candidate control instruction based on the status monitoring data and operating condition data, and sends the farm-level candidate control instruction to the unit edge control device 11; the unit edge control device 11 analyzes the rationality of the candidate instructions based on the single-unit candidate control instruction and the farm-level candidate control instruction, thereby determining a reasonable target control instruction and sending it to the wind turbine control device 12, so that the wind turbine control device 12 drives the actuator 15 to perform the action indicated by the target control instruction. The wind turbine control device 12 does not unconditionally execute the field-level control instructions issued by the field-level control system 13. Instead, it analyzes the rationality of the field-level candidate control instructions through the unit edge control device 11 to determine the target control instruction to be executed by the wind turbine. This avoids the wind turbine from executing abnormal control instructions generated by the field-level control system 13 due to software defects, hardware failures, model drift, misconfiguration, bad data, or other anomalies. In this way, it prevents the actuator 15 from performing out-of-bounds actions, ensuring the normal operation of the wind turbine and improving the safety of wind farm control.

[0044] In some embodiments, in order to improve the security isolation between the unit edge control device 11 and the field-level control system 13, a data isolation communication device may be provided between the unit edge control device 11 and the field-level control system 13. Figure 3 This is a schematic diagram of the structure of a wind farm sidefield coordinated control system provided in another embodiment of this application. Figure 3 and Figure 2 The difference is that, Figure 3 The wind farm side-field coordinated control system shown may also include a data isolation communication device 14.

[0045] The data isolation communication device 14 connects the turbine edge control device 11 and the field-level control system 13, meaning that the turbine edge control device 11 and the field-level control system 13 communicate via the data isolation communication device 14. The data isolation communication device 14 includes a first data channel 141 and a second data channel 142. The first data channel 141 is used by the turbine edge control device 11 to transmit data to the field-level control system 13; the first data channel 141 can be considered an uplink data channel. The turbine edge control device 11 can transmit status monitoring data and operating condition data to the field-level control system 13 through the first data channel 141. In some examples, the execution feedback after the wind turbine executes the target control instruction can also be reflected through status monitoring data and operating condition data. The second data channel 142 is used by the field-level control system 13 to transmit data to the turbine edge control device 11; the second data channel 142 can be considered a downlink data channel. The field-level control system 13 can transmit field-level candidate control instructions to the turbine edge control device 11 through the second data channel 142. In some examples, the field-level control system 13 can also transmit security policy versions and unit scheduling constraints to the unit edge control device 11 via the second data channel 142. In some examples, a verification module 143 can be configured for the second data channel 142. The verification module 143 verifies the field-level candidate control indications transmitted through the second data channel 142 to obtain an indication verification result, which indicates whether the field-level candidate control indication has passed verification. Verification of the field-level candidate control indications may include, but is not limited to, one or more of the following: indication whitelist verification, signature verification, serial number verification, timestamp verification, and log auditing. Indicator whitelist verification verifies whether the field-level candidate control indication is within a pre-configured indication whitelist; signature verification verifies the signature of the field-level candidate control indication; serial number verification verifies the serial number of the field-level candidate control indication; timestamp verification verifies the timestamp of the field-level candidate control indication; and log auditing verifies the log that generated the field-level candidate control indication. In some examples, the first data channel 141 can be implemented by one or more of the following: a unidirectional optical shutter, a unidirectional data diode, a data gateway, an industrial firewall, a message queue isolation gateway, etc., and is not limited here; the second data channel 142 can be implemented by one or more of the following: an independent control gateway, a whitelist control channel, a signature control bus, a virtual private network (VPN), etc., and is not limited here.

[0046] The data isolation communication device 14 enables network isolation for wind farm edge-field coordinated control. The wind farm's control closed loop is disconnected at the network boundary, i.e., at the data isolation communication device 14. However, through rationality analysis of the unit edge control device 11 and determination of target control instructions, the control closed loop is re-closed locally at the wind turbine side, i.e., at the unit edge control device 11. Decision-making, execution, and feedback are completed at the wind turbine side, eliminating time delay and reliability limitations imposed by cross-network links. This balances farm-level closed-loop control with isolation between the wind turbine side and the farm-level control system 13. The first data channel can employ a unidirectional data diode or a unidirectional data shutter. In this embodiment, the control closed loop no longer relies on cross-boundary bidirectional real-time control, thus isolating farm-level anomaly indications from being transmitted to the wind turbine. Especially when the wind turbine side and the farm-level control system 13 are located in different safety zones—for example, the wind turbine side includes the turbine edge control device 11, the wind turbine control device 12, and the actuator 15, with the wind turbine side located in the production control area and the farm-level control system 13 located in the management information area—it is even more necessary to install a data isolation communication device 14 between the wind turbine side and the farm-level control system 13 to ensure safety isolation between the production control area and the management information area. The turbine edge control device 11 determines the target control instruction, which can effectively curb out-of-bounds instructions caused by anomalies in the farm-level control system 13 such as software defects, hardware failures, model drift, misconfiguration, bad data, miscalculations, communication damage, and intrusion, thus preventing the wind turbine from executing out-of-bounds instructions.

[0047] Experiments verified that in a 500MW wind farm, each of the 50 10MW wind turbines was equipped with a unit edge control device 11, and a farm-level control system 13 was configured at the substation. The first data channel 141 could be a one-way optical shutter. Each unit edge control device 11 uploaded status monitoring data and operating data to the farm-level control system 13 through the first data channel 141, and injected out-of-bounds or erroneous farm-level candidate control instructions into the farm-level control system 13. All out-of-bounds or erroneous farm-level candidate control instructions were rejected by the unit edge control devices 11, effectively curbing out-of-bounds instructions and improving the safety of wind turbines in the wind farm.

[0048] In some embodiments, the unit edge control device 11 may include a first control module 111 and a second control module 112. The first control module 111 is a conventional control module used for routine control, such as acquiring status monitoring data and operating condition data, performing preprocessing such as data cleaning, generating single-unit candidate control instructions, and determining target control instructions. Specifically, the first control module 111 can be used to generate single-unit candidate control instructions based on status monitoring data, operating condition data, and single-unit optimized control objectives, and to obtain target control instructions based on single-unit candidate control instructions and field-level candidate control instructions. The single-unit optimized control objective is the optimization objective in the single-unit control optimization logic of the above embodiments, as described above, and will not be repeated here. The first control module 111 can analyze the field-level candidate control instructions to determine their rationality. If the field-level candidate control instructions are rational, they can be used to generate the target control instruction; if the field-level candidate control instructions are unreasonable, they are rejected, and the single-unit candidate control instructions or other local control instructions of the wind turbine are used to generate the target control instruction. In some examples, such as... Figure 2 As shown, the first control module 111 may include a data acquisition unit 1111, a single-machine optimization unit 1112, and an indication determination unit 1113; the data acquisition unit 1111 can be used to receive status monitoring data and operating condition data; the single-machine optimization unit 1112 is used to generate single-machine candidate control indications based on the status monitoring data and operating condition data; the indication determination unit 1113 is used to determine the target control indication based on the single-machine candidate control indications and the field-level candidate control indications.

[0049] The second control module 112 is a safety control module independent of the first control module 111, used for safety control. Even if the first control module 111 fails or the communication network fails, the second control module 112 can still independently control the wind turbine control device or actuator. An independent control channel 18 is provided between the second control module 112 and the wind turbine control device 12, or an independent control channel is provided between the second control module 112 and the actuator 15. Figure 2 Taking the example of an independent control channel between the second control module 112 and the wind turbine control device 12, therefore Figure 2(The independent control channel between the second control module 112 and the actuator 15 is not shown in the diagram). The independent control channel 18 is a communication channel independent of the first data channel 141 and the second data channel 142, and can be used to transmit instructions for shutdown, load reduction, safety limiting, etc., in emergency situations. The existence of the independent control channel 18 ensures that even if the first control module 111 fails or communication between the first control module 111 and other structures fails, the second control module 112 can independently control the wind turbine control device 12 or the actuator 15 to ensure the safety of the wind turbine. When a first predetermined condition is triggered, the second control module 112 transmits a control instruction for the first predetermined condition through the independent control channel 18, so that the wind turbine control device 12 or the actuator 15 executes the control instruction. The first predetermined condition is used to determine the occurrence of an emergency safety event. That is, the first predetermined condition is the condition for determining the occurrence of an emergency safety event. When the first predetermined condition is triggered, it indicates that an emergency safety event has occurred. The second control module 112 can be directly triggered to generate a control instruction for the first predetermined condition. This control instruction for the first predetermined condition is the control instruction for the emergency safety event. The control instruction for the first predetermined condition may include, but is not limited to, a shutdown instruction, a load reduction instruction, etc. This control instruction is transmitted through the independent control channel 18 to perform timely safety control and protect the wind turbine's safety. It should be noted that if the wind turbine control device 12 receives a target control instruction sent by the first control module 111 and a control instruction for the first predetermined condition transmitted by the second control module 112 through the independent control channel within the same time period, the wind turbine control device 12 executes the control instruction for the first predetermined condition. The control instruction for the first predetermined condition has a higher priority than the target control instruction.

[0050] In some examples, the independent control channel 18 can be one or more of the following: hardwired safety chain, Controller Area Network (CAN) bus, functional safety PLC channel, safety relay channel, etc. A completely independent hardware structure can be configured for the second control module 112, for example, Figure 4 A schematic diagram illustrating an example of a unit edge control device provided in an embodiment of this application, as shown below. Figure 4As shown, the wind turbine edge control device 11 includes a first control module 111 and a second control module 112. The second control module 112 includes an independent microcontroller unit (MCU) 1121, an independent memory 1122, an independent watchdog timer 1123, an independent power supply 1124, an independent clock 1125, an independent shutdown channel 1126, an independent load reduction channel 1127, and an independent core 1128. The independent shutdown channel 1126 and the independent load reduction channel 1127 are directly connected to the wind turbine control device 12. The independent hardware structure of the second control module 112 allows it to send shutdown instructions, load reduction instructions, etc., to the wind turbine control device 12 or the actuator 15 independently, without relying on the first control module 111 or the conventional Ethernet link.

[0051] For example, Figure 5 This is a schematic diagram illustrating an example of the interaction between the turbine edge control device and the wind turbine control device provided in an embodiment of this application. Figure 5 As shown, the first control module 111 in the unit edge control device 11 can be connected to the wind turbine control device 12 via conventional communication paths such as Ethernet and control bus. The first control module 111 and the second control module 112 can also be connected via conventional communication paths such as Ethernet and control bus. The second control module 112 is connected to the wind turbine control device 12 via an independent control channel 18, which is independent of the conventional communication path between the first control module 111 and the wind turbine control device 12. The wind turbine control device 12 may include a first input terminal 121 and a second input terminal 122; the first input terminal 121 is used to connect to the first control module 111 and to the actuator 15; the second input terminal 122 is connected to the second control module 112 through an independent control channel 18, and the second input terminal 122 may also be connected to the actuator 15. The priority of the second input terminal 122 is higher than that of the first input terminal 121. The priority of the control instructions transmitted between the second input terminal 122 and the actuator 15 is higher than that of other control instructions. The control instructions transmitted from the second input terminal 122 to the actuator 15 may cover the control instructions transmitted through other conventional communication paths.

[0052] In some embodiments, the unit edge control device 11 can determine the target control instruction with the assistance of safety control auxiliary data. The unit edge control device 11 can acquire safety control auxiliary data and, based on individual unit candidate control instructions, safety control auxiliary data, and farm-level candidate control instructions, determine the target control instruction to be allocated to the wind turbine control device 12. The safety control auxiliary data is used to assist in determining the relative importance of individual unit candidate control instructions and farm-level candidate control instructions; in other words, the safety control auxiliary data is used to assist in determining the rationality of farm-level candidate control instructions. The rationality of farm-level candidate control instructions is related to the relative importance of individual unit candidate control instructions and farm-level candidate control instructions. The higher the rationality of farm-level candidate control instructions, the higher their importance, and correspondingly, the higher their weight in determining the target control instruction. If a farm-level candidate control instruction is not rational, such as meeting the degradation control conditions described below, the unit edge control device 11 can reject the farm-level candidate control instruction and obtain the target control instruction using individual unit candidate control instructions or other local wind turbine instructions.

[0053] In some examples, safety control auxiliary data may include factors that can influence the importance of individual unit candidate control indications and farm-level candidate control indications in wind turbine control. Safety control auxiliary data may include, but is not limited to, at least one of the following: safety status, communication status, individual unit preset operating boundaries, individual unit safety envelope, indication verification results, channel integrity monitoring results, and reliability data. Safety status is used to characterize whether a routine safety event has been triggered; for example, if the temperature and vibration of the wind turbine's main shaft bearing are abnormal, triggering a main shaft bearing safety event, the safety status can characterize the triggering of the main shaft bearing safety event; conversely, if no routine safety event occurs in the wind turbine, the safety status can characterize that the wind turbine is operating normally and no routine safety event has been triggered. The communication status can be used to characterize the communication status between the unit edge control device 11 and the field-level control system 13, and / or the communication status between the second control module 112 in the unit edge control device 11 and the wind turbine control device 12; the communication status can be obtained by monitoring the heartbeat signal between the unit edge control device 11 and the field-level control system 13; the communication status can be obtained by monitoring the heartbeat signal between the second control module 112 and the wind turbine control device 12; the communication status may include, but is not limited to, normal communication, communication interruption, and communication delay; it can be obtained through... Figure 2The communication detection unit 1114 in the turbine edge control device 11 performs monitoring operations to obtain the communication status. The single-unit preset operating boundary is the boundary of the operating data that the wind turbine must comply with, which can be determined by the wind turbine's own capabilities. The single-unit safety envelope (which can be called the Local Safety Envelope) is the dynamic safety boundary of the wind turbine's operating data obtained based on condition monitoring data, operating condition data, and the lightweight safety simulation sub-model in the turbine edge control device. The single-unit safety envelope can be a feasible control domain jointly defined by multiple constraints; for example, the single-unit safety envelope can be a feasible control domain jointly defined by multiple constraints such as load, temperature, speed, vibration, pitch angle, and yaw angle. The indication verification result is the verification result of the field-level candidate control indication, which can be obtained through the operation of the verification module 143 in the data isolation communication device 14. For details, please refer to the relevant descriptions in the above embodiments, which will not be repeated here. The channel integrity monitoring results are used to characterize whether the overlapping data in the data channel of the data isolation communication device 14 and the independent control channel 18 diverges. The data channels include the first data channel 141 and the second data channel 142. The first data channel 141, the second data channel 142 and the independent control channel 18 may transmit the same type of data, i.e., there is overlapping data. If there is overlapping data, the same type of data transmitted in the first data channel 141, the second data channel 142 and the independent control channel 18 can be compared to determine whether the same type of data transmitted in different channels diverges, i.e., whether the overlapping data transmitted in different channels diverges. Here, divergence can refer to the difference between the values ​​of the same type of data transmitted in different channels being greater than the difference threshold, or it can refer to the difference between the values ​​of the same type of data transmitted in different channels continuously increasing over time. Channel integrity monitoring results can be obtained through the operation of the cross-channel module 144 in the data isolation communication device 14. In one or more embodiments, the gap threshold is a threshold value used to determine whether the same type of data is diverging when transmitted in different channels. The gap threshold can be determined based on the error limits of the same type of data in different channels and the dynamic error caused by the allowable time alignment deviation, provided that the data type and the time of data acquisition are aligned respectively. For example, if the error limits of the same type of data in two channels are eA,q and eB,q respectively, the allowable time alignment deviation is δ, and the allowable maximum change rate of the data of this type under the current operating mode is Rq, then the comparison error limit Eq=eA,q+eB,q+Rq×δt, and the gap threshold θ(G,q)=max[Q0.99(Gq,normal),Eq], where Q0.99 is the 99th percentile. Rq is taken as the upper limit of the change rate of the unit control configuration or the verified normal response; δt is taken as the upper limit of the allowable error of time synchronization and sampling alignment. Data with a time alignment deviation exceeding δt is not subject to effective numerical comparison and is treated as unusable data transmitted in the channel.The gap threshold θ(G,q) should also be less than the maximum permissible cross-channel deviation specified by control and safety requirements for that type of data. Confidence data is used to characterize the confidence of any one or more of the data acquisition channels, operating condition data, sub-models, and field-level candidate control indications of the data acquisition device. The data acquisition channel is the acquisition channel through which the unit edge control device 11 obtains status monitoring data. For example, the data acquisition channel may include the channel through which sensors in the multi-source sensor group 16 in the above embodiment transmit data to the unit edge control device 11, but is not limited to this. The sub-model may include a lightweight safety simulation sub-model and multiple physical sub-models related to the wind farm and wind turbine set in the field-level control system. The lightweight safety simulation sub-model is a lightweight quantum model that resides locally on the wind turbine and can run in real time, participating in the determination of the safety state and the determination of the single-unit safety envelope; the lightweight safety simulation sub-model does not rely on the wind farm's field-level computing power and cross-network communication links. The lightweight safety simulation sub-model may include, but is not limited to, one or more of the following: reduced-order load sub-model, temperature rise sub-model, and envelope generation sub-model. The reduced-order load sub-model can estimate the equivalent fatigue load or ultimate load of the wind turbine's load-concerned structure in real time based on operating condition data such as speed, pitch angle, and power. In some examples, the reduced-order load sub-model can be an internal model built based on measurement data from blade root load sensors. Three fiber optic load sensors can be installed at the root of each blade, and these three fiber optic load sensors can be arranged circumferentially along the blade root cross-section. The wavelength signals collected by the fiber optic load sensors are demodulated by fiber optics, wavelength-to-strain conversion, zero-point compensation, and temperature compensation to obtain strain data, and then... The installation positions, angles, and load calibration matrices of the three fiber optic load sensors are used to calculate the bending moments and combined loads of the blades in the flapping and oscillation directions. During model construction, time synchronization, outlier removal, and operating condition partitioning can be performed on blade root load data and operating condition data such as rotational speed, impeller azimuth angle, pitch angle, power, torque, and wind speed. One or more of the following methods—response surface methodology, regression model, state-space identification, neural network, or high-fidelity model reduction—can be used to establish a mapping relationship between operating condition data and equivalent fatigue loads, ultimate loads, or load change trends. Parameters required for constructing the reduced-order load sub-model may include the installation positions and angles of the fiber optic load sensors, wavelength-strain conversion coefficients, zero-point compensation coefficients, temperature compensation coefficients, load calibration matrix, sampling period, statistical window, effective sector of impeller azimuth angle, model order or model weights, and the first, second, and third load thresholds determined based on the DLC13 wind turbine design load conditions and load reduction requirements. Different load thresholds correspond to different fiber optic load sensors or blades, and are not limited here.The temperature rise sub-model can predict the temperature change trend of the wind turbine's temperature-related locations based on factors such as heat dissipation power, heat dissipation conditions, and temperature data. It can predict short-term or long-term trends. In some examples, the temperature rise sub-model can be an internal model built based on temperature monitoring data and operating condition data. Temperature monitoring data may include, but is not limited to, at least one of the following: first pitch motor temperature, second pitch motor temperature, third pitch motor temperature, generator winding temperature, maximum generator winding temperature, generator front bearing inner ring temperature, generator front bearing outer ring temperature, generator rear bearing inner ring temperature, and generator rear bearing outer ring temperature. The first, second, and third pitch motor temperatures each correspond to the pitch motor of the blade. If the wind turbine has more blades, it may have more pitch motor temperatures. Operating condition data may include, but is not limited to, power, speed, torque, current, pitch angle, and pitch operation status. The model requires at least one of the following: ambient temperature, cabin temperature, cooling system operating status, and wind speed. During model construction, temperature monitoring data and operating condition data can be synchronized in time, outlier removal can be performed, and operating condition partitions can be established. The current temperature, temperature change rate, or predicted temperature at the temperature of interest location can be used as the output. A single-node or multi-node thermal resistance-thermal capacity network, state-space identification, regression model, or neural network can be employed to establish a mapping relationship between operating condition data and temperature rise. Historical operating data or experimental data can be used for parameter identification and model verification. The parameters required for constructing the temperature rise sub-model may include, but are not limited to, at least one of the following: temperature measurement point location, initial temperature, sampling period, prediction time domain, thermal resistance, thermal capacity, heat source loss coefficient, heat dissipation coefficient, operating condition feature weights, time lag parameters, and temperature warning threshold. The temperature rise sub-model can output the short-term or long-term temperature change trend at the temperature of interest location, the predicted maximum temperature, and the estimated time to reach the temperature warning threshold.The envelope generation sub-model can dynamically generate a single-machine safety envelope based on state monitoring data, operating condition data, and the single-machine preset operating boundary. In some examples, the envelope generation sub-model can be built using commercial modeling and optimization platforms such as MATLAB / Simulink and its Model Predictive Control Toolbox, or it can be built using an internal constraint aggregation model. The internal constraint aggregation model can take the load prediction results output by the reduced-order load sub-model, the temperature prediction results output by the temperature rise sub-model, and the constraints such as speed, vibration, pitch angle, power, and yaw angle in the single-machine preset operating boundary as envelope constraints. Based on the current state of the wind turbine and candidate control indications, the corresponding envelope constraints are predicted. The system generates a single-machine safety envelope by intersecting the set of control variables that satisfy all envelope constraints with the operating variables. This single-machine safety envelope can be represented using one or more forms, such as a polyhedron, ellipsoid, piecewise linear region, or discrete lookup table. The parameters required to construct the envelope generation sub-model may include, but are not limited to, at least one of the following: upper and lower limits for each operating variable, load threshold, temperature threshold, model coefficients, prediction time domain, control variable change rate limit, safety margin, confidence margin, and envelope update period. If no feasible control variable satisfies all envelope constraints, the envelope generation sub-model can output results such as rejecting field-level candidate control indications, adopting safety limiting, or adopting single-machine candidate control indications. A lightweight safety simulation sub-model can be used. Figure 2 The lightweight model unit 1115 is implemented in the code. It can be accessed via... Figure 2 The operation of the credibility assessment unit 1116 in the system obtains credibility data. In some examples, the credibility assessment unit 1116 can first perform data cleaning on the condition monitoring data and operating condition data, and then generate credibility data of the cleaned condition monitoring data and operating condition data.

[0054] In one or more embodiments, for the thresholds involved in the embodiments of this application, a parameter table can be established according to the aircraft type, operating mode, and the data being compared, recording at least the data unit, judgment period Ts, comparison time interval ΔT, statistical window, data validity period, error limit, and each threshold. Normal steady-state, start-stop, pitch regulation, and power ramp operating modes each use the parameter group corresponding to their respective operating modes, and normal ramp samples are not mixed with steady-state samples to determine the threshold. When switching operating modes, the corresponding statistical window is refilled. If the statistical window is not full or there are no valid parameters for that operating mode, field-level optimized control is not restored accordingly. The threshold is fixed within an operating phase. If abnormal samples exist, these abnormal samples will not be used online to relax the threshold. Normal operating condition data that has passed sensor validity, time synchronization, communication verification, and unit operating boundary checks can be acquired. At least 1000 non-overlapping statistical windows are formed using the same data processing method as online data as calibration samples; separately, data not involved in calibration is reserved for verification. For a sample set Z containing n values, the values ​​are sorted in ascending order. The empirical quantile Qp(Z) is defined as the ceil(p×n)th value after sorting, where 0 < p ≤ 1, and ceil represents rounding up. Q0.99 can be used as the upper limit of normal variation or difference, i.e., the upper limit type threshold, and Q0.01 can be used as the lower limit of normal variation or difference, i.e., the lower limit type threshold. Q0.99 is the 99th quantile, and Q0.01 is the 1st quantile.

[0055] In the above embodiments, the lightweight safety simulation sub-model configured on the edge control device 11 of the wind turbine and the physical sub-model configured on the field-level control system 13 can jointly form a federated or centralized and distributed hybrid mode of the edge field model. The physical sub-model can be divided according to its responsibilities such as computing power, latency, and security, which is not limited here. The lightweight safety simulation sub-model and the physical sub-model can be coupled through data transmission via the first data channel 141 and control instruction transmission via the second data channel 142. The lightweight safety simulation sub-model has the characteristics of low latency and does not rely on cross-network communication links and field-level computing power. The lightweight safety simulation sub-model runs locally on the wind turbine in real time to ensure the local safety baseline. The physical sub-model has the characteristics of high fidelity and runs on the field-level control system 13 to serve the overall optimization, realizing the collaborative closed loop of safe autonomy of the wind turbine and overall optimization of the wind farm.

[0056] The credibility data of any object can be obtained by fusing multiple dimensions of features of that data or indication. In some examples, for the credibility data of objects such as data acquisition channels, operating condition data, and sub-models, the credibility assessment unit 1116 can obtain the credibility data of such objects based on their statistical relevance, anomalous content / anomaly suppression operations, completeness, timeliness, and stability. The anomalous content / anomaly suppression operations can be obtained through unsupervised anomaly detection; anomalous content may include anomalous points and / or anomalous fragments, and anomaly suppression operations are operations that suppress anomalous content. The credibility data of such objects can be obtained by fusing their statistical relevance, anomalous content, completeness, timeliness, and stability. The fusion method may include, but is not limited to, one or more of the following: weighted summation, product calculation, minimum value calculation, Bayesian fusion, DS evidence fusion, and neural network fusion.

[0057] For the credibility data of field-level candidate control indications, the credibility assessment unit 1116 can obtain the credibility data of field-level candidate control indications based on their source, signature, scope, timeliness, and consistency with the state of the wind turbine. The credibility data of field-level candidate control indications can be obtained by fusing the source, signature, scope, timeliness, and consistency with the state of the wind turbine. The fusion algorithm may include, but is not limited to, one or more of the following: weighted summation, product calculation, minimum value calculation, Bayesian fusion, DS evidence fusion, and neural network fusion.

[0058] For example, Figure 6 A schematic diagram illustrating an example of the process for determining credibility data provided in embodiments of this application, such as... Figure 6 As shown, the reliability assessment unit 1116 may include a first assessment subunit 21, a second assessment subunit 22, a third assessment subunit 23, a fourth assessment subunit 24, a fifth assessment subunit 25, and a sixth assessment subunit 26. Input objects may include data acquisition channels, operating condition data, sub-models, and field-level candidate control indications. Specifically, the data acquisition channels, operating condition data, and sub-models are input into the first assessment subunit 21, second assessment subunit 22, third assessment subunit 23, fourth assessment subunit 24, and fifth assessment subunit 25, respectively, while the field-level candidate control indications are input into the sixth assessment subunit 26. The first assessment subunit 21 can be used to perform statistical correlation analysis on the input objects such as data acquisition channels, operating condition data, and sub-models to obtain the statistical correlation α of the input objects. j The second evaluation subunit 22 can be used to perform unsupervised anomaly detection on input objects such as data acquisition channels, operating condition data, and sub-models, detecting data points or data fragments that deviate from normal behavior patterns, thereby obtaining abnormal content / anomaly suppression operation b. jThe third evaluation subunit 23 can be used to analyze the integrity of input objects such as data acquisition channels, operating condition data, and sub-models to obtain the integrity g of the input objects. j The fourth evaluation subunit 24 can be used to analyze the timeliness of input objects such as data acquisition channels, operating condition data, and sub-models, and obtain the timeliness h of the input objects. j The fifth evaluation subunit 25 can be used to analyze the stability of input objects such as data acquisition channels, operating condition data, and sub-models, and obtain the stability s of the input objects. j By analyzing the statistical correlation a j Abnormal content / abnormal suppression operation b j Integrity g j Timeliness j and stability j The fusion of these data sources yields credibility data for input objects such as data acquisition channels, operating condition data, and sub-models. For example, credibility data dj=F(a j b j g j h j s j F() represents the fusion algorithm. The sixth evaluation subunit 26 can be used to fuse the source, signature, range, timeliness, and consistency with the state of the wind turbine at the field level candidate control indication, to obtain the credibility data c of the field level candidate control indication. j The credibility vector can be formed by combining the credibility data of input objects such as data acquisition channels, operating condition data, sub-models, and field-level candidate control indications. Each element in the credibility vector represents the credibility data of each input object. The credibility vector can be provided to the unit edge control device 11 to assist it in generating single-unit candidate control indications and determining target control indications. For example, the unit edge control device 11 can generate single-unit candidate control indications based on single-unit mirror state, wind speed, wind direction, load, and credibility data. The single-unit mirror state can be obtained based on the lightweight safety simulation sub-model configured in the unit edge control device 11. The credibility vector can also be provided to the field-level control system 13 to assist it in generating field-level candidate control indications. For example, it can reduce the influence weight of wind turbines and input objects with low credibility data on the generation of field-level candidate control indications. The confidence vector can be provided to the field-level control system 13 to assist in updating the model parameters in the physical sub-model involved in the field-level control system 13. For example, it can reduce the gain of updating the model parameters in the physical sub-model for input objects with low confidence represented by confidence data.

[0059] In some examples, condition monitoring data, operating condition data, and a confidence vector including confidence data can be input into multiple physical sub-models. This yields the output of each physical sub-model based on the condition monitoring data and operating condition data. These outputs can then be weighted and normalized according to the confidence data to obtain a fused output. The residual between the fused output and the actual data is calculated, and the model parameters of the physical sub-models are updated with gains based on the confidence data and the residual. For example, Figure 7 A schematic diagram illustrating an example of the gain for determining the model parameter update of a physical sub-model provided in an embodiment of this application, as shown below. Figure 7 As shown, the cleaned state monitoring data and operating condition data X', along with the confidence vector, can be input into the aerodynamic sub-model, transmission chain sub-model, electromagnetic sub-model, and tower structure sub-model, respectively. The aerodynamic sub-model outputs Y_1, the transmission chain model outputs Y_2, the electromagnetic sub-model outputs Y_3, and the tower structure sub-model outputs Y_4. The confidence data can be normalized and weighted to obtain the fused output Y. The residual e between the fused output Y and the actual data Y_real is calculated. The gain is updated based on the confidence data and the residual. The fused output Y, residual, and gain can be obtained according to the following formulas (1) to (4): Y = Σ(α_i × Y_i) (1) α_i = φ(t_i) / [Σφ(t_i) + ε] (2) e = Y_real – Y (3) θ_i(k+1) = θ_i(k) + K0 × φ(t_i) × e_i (4) Where Y is the fusion output; Y_i is the output of the i-th physical sub-model; α_i is the fusion weight corresponding to the output of the i-th physical sub-model; t_i is the confidence data of the input data of the i-th physical sub-model; φ() represents the mapping function from confidence data to weights; ε is the smoothing factor; e is the residual; Y_real is the actual data; θ_i(k+1) is the gain obtained in the (k+1)-th update; θ_i(k) is the gain obtained in the k-th update; K0 is the learning coefficient; e_i is the residual of the i-th physical sub-model.

[0060] Experimental verification showed that in a 10MW doubly-fed induction generator (DFIG) wind turbine scenario, the unit's edge control device 11 received data from 32 sensors, including 11 vibration data points, 8 temperature data points, 9 strain data points, and 6 electrical data points. The reliability data determination period for the sensor data was 100ms, and the time-series feature compression window was divided into 16 segments. The statistical correlation threshold ranged from 0.5 to 0.8. Unsupervised anomaly detection employed an isolated forest model, a Local Outlier Factor (LOF) model, or an autoencoder, without limitation. Experimental verification demonstrated that the introduction of reliability data reduced the impact of outlier data on updating the physical sub-model, resulting in a decrease in the residuals and false alarm rate of the physical sub-model, thus improving its accuracy.

[0061] In some embodiments, the first data channel 141 can classify the transmitted data by traffic, prioritizing the transmission of high-priority data. For example, data related to security events has the highest priority, and the first data channel 141 prioritizes transmitting such data when it exists. The data transmission triggering method of the first data channel 141 may include, but is not limited to, one or more of the following: periodic triggering, event triggering, change triggering, and reliability triggering. Periodic triggering means data transmission is triggered according to a cycle; event triggering means data transmission is triggered when an event occurs; change triggering means data transmission is triggered when the change in data reaches the change triggering condition; and reliability triggering means data transmission is triggered when the reliability of the data reaches the reliability triggering condition. In some examples, when a unit decision triggering event occurs, the first data channel 141 changes the data transmission parameters. Unit decision triggering events are important events of concern to wind turbines and wind farms, indicating that the state of the wind turbine exceeds the normal range. When a unit decision triggering event occurs, the farm-level control system 13 requires high-resolution data for analysis and processing. Correspondingly, the first data channel can change the data transmission parameters to enable the first data channel 141 to transmit high-resolution data. Data transmission parameters can characterize the resolution of the transmitted data. For example, data transmission parameters may include, but are not limited to, data transmission frequency. When data transmission parameters include data transmission frequency, the data transmission frequency can be increased, which allows the field-level control system 13 to acquire more detailed data. In some examples, unit decision-triggered events may include, but are not limited to, one or more of the following: safety events, changes in operating condition data exceeding a first change threshold, changes in condition monitoring data exceeding a second change threshold, and changes in the reliability represented by reliability data exceeding a third change threshold. Safety events may include routine safety events and emergency safety events. Emergency safety events are more severe and have a higher priority than routine safety events. The first change threshold is a threshold for changes in operating condition data that may represent abnormal changes. In one or more embodiments, the first change threshold can be determined based on the statistical upper limit of the absolute value of the change in operating condition data obtained under normal operating conditions using the same comparison time interval and the error limit of the operating condition data. The second change threshold is a threshold for the change in condition monitoring data that may indicate an anomaly. In one or more embodiments, the second change threshold can be determined based on the statistical upper limit of the absolute value of the change in condition monitoring data obtained under normal operating conditions using the same comparison time interval and the error limit of the condition monitoring data. The third change threshold is a threshold for the confidence level that may indicate an anomaly. In one or more embodiments, the third change threshold can be determined based on the statistical upper limit of the absolute value of the change in confidence level obtained under normal operating conditions using the same comparison time interval and the error limit of the confidence level.For example, for each operating condition quantity q, a first change threshold θ1,q=max[Q0.99(abs(Δxq) normal),2ex,q] is determined; for each state monitoring quantity q, a second change threshold θ2,q=max[Q0.99(abs(Δyq) normal),2ey,q] is determined; and for the compared reliability, a third change threshold θ3=max[Q0.99(abs(ΔC) normal),2εC] is determined. ex,q and ey,q are the error limits of the corresponding data, εC is the normal fluctuation tolerance of the reliability score, and the change in operating condition data Δxq(k)=xq(k). xq( m) is the difference in operating condition data before and after the comparison time interval, and the change in condition monitoring data is Δyq(k) = yq(k). yq( m) is the difference between the state monitoring data before and after the comparison time interval, and the change in reliability is ΔC(k) = C(k). C( m) is the difference in credibility before and after the comparison time interval. In other cases, the first change threshold, the second change threshold, and the third change threshold can also be set according to specific scenarios, needs, experience, etc., and are not limited here. High-resolution data is transmitted only through the first data channel 141 when a unit decision-triggered event occurs; in other cases, low-resolution data can be transmitted to adapt to the throughput limit of the unidirectional first data channel and reduce bandwidth consumption.

[0062] In some embodiments, degradation control conditions can be preset, and the unit edge control device 11 can use these conditions to determine whether the target control indication should adopt a field-level candidate control indication. The unit edge control device 11 can be used to: generate a target control indication based on a field-level candidate control indication when the safety control auxiliary data and the field-level candidate control indication do not meet the degradation control conditions; and reject a field-level candidate control indication when the safety control auxiliary data or the field-level candidate control indication meets the degradation control conditions, and determine the target control indication based on the met degradation control conditions. Degradation control conditions include conditions where the field-level candidate control indication is unreasonable and should be rejected. If the safety control auxiliary data and the field-level candidate control indication do not meet the degradation control conditions, it indicates that the field-level candidate control indication is reasonable, and the target control indication can be determined using the field-level candidate control indication. In this case, the target control indication is a field-level candidate control indication, or the target control indication is formed by fusing the field-level candidate control indication with a single-unit candidate control indication. When the target control indication is a field-level candidate control indication, its usage weight is 100%. When the target control indication is a fusion of field-level and single-unit candidate control indications, both the usage weights of the field-level and single-unit candidate control indications are greater than zero, and their respective usage weights can be determined based on specific circumstances. If safety control auxiliary data or field-level candidate control indications meet the degradation control conditions, it indicates that the field-level candidate control indication is unreasonable and should be rejected. Correspondingly, the usage weight of the field-level candidate control indication is 0%, and the usage weight of the single-unit candidate control indication can be in the range of 0% to 100%. In this case, the target control indication may include, but is not limited to, one or a fusion of two or more of the following: single-unit candidate control indication, load reduction indication, safety limit indication, unit default control curve, and control indication corresponding to points within the projection range of the single-unit safety envelope.

[0063] In some examples, degradation control conditions may include, but are not limited to, one or more of the following: the security status representation in the security control auxiliary data triggers a regular security event; the integrity monitoring representation in the security control auxiliary data shows overlapping data divergence, communication status representation communication interruption, or indication verification result representation indicating failure to verify the field-level candidate control indication; any control quantity in the field-level candidate control indication exceeds the single-machine security envelope in the security control auxiliary data; or the credibility of the credibility data representation in the security control auxiliary data is lower than the credibility threshold.

[0064] Degradation control conditions may include a routine safety event triggered by a safety status representation in the safety control auxiliary data. When a routine safety event is triggered by a safety status representation in the safety control auxiliary data, the unit edge control device 11 rejects the field-level candidate control instruction. The target control instruction determined by the unit edge control device 11 may include a shutdown instruction, a load reduction instruction, or a safety limit instruction, in order to mitigate or suspend the adverse effects of the safety event on the wind turbine and wind farm as quickly as possible.

[0065] Degradation control conditions may include: channel integrity monitoring results in the safety control auxiliary data indicating data divergence overlapping between the data channels in the data isolation communication device and the independent safety control channel; communication status indicating communication interruption; or indication verification results indicating failure to verify the field-level candidate control indication. Channel integrity monitoring results indicating data divergence overlapping between the data channels in the data isolation communication device and the independent safety control channel can determine an anomaly in the control surface. Communication status indicating communication interruption means that the unit edge control device 11 may not be able to obtain the field-level candidate control indication in a timely manner, or the field-level candidate control indication obtained by the unit edge control device 11 is unreliable. Indication verification results indicating failure to verify the field-level candidate control indication indicate that the field-level candidate control indication is unreliable. In the event of channel integrity monitoring results in the safety control auxiliary data indicating data divergence overlapping between the data channels in the data isolation communication device and the independent safety control channel, communication status indicating communication interruption, or indication verification results indicating failure to verify the field-level candidate control indication, the unit edge control device 11 rejects the field-level candidate control indication. The target control indication determined by the unit edge control device 11 may include a single-unit candidate control indication or a unit default control curve to achieve single-unit autonomous control of the wind turbine.

[0066] Degradation control conditions may include anomalies in the correlation representation between different types of status monitoring data and operating condition data. For example, anomalies may be determined through methods such as redundant voting and cross-validation among different types of status monitoring data and operating condition data. When anomalies are found in the correlation representation between different types of status monitoring data and operating condition data, the unit edge control device 11 rejects the field-level candidate control instruction. The target control instruction determined by the unit edge control device 11 may include a single-unit candidate control instruction or the unit's default control curve to achieve single-unit autonomous control of the wind turbine.

[0067] Degradation control conditions may include any control quantity in the field-level candidate control indications exceeding the single-machine safety envelope in the safety control auxiliary data. If any control quantity in the field-level candidate control indications exceeds the single-machine safety envelope in the safety control auxiliary data, it indicates that the field-level candidate control indication is an out-of-bounds indication. When any control quantity in the field-level candidate control indications exceeds the single-machine safety envelope in the safety control auxiliary data, the field-level candidate indication can be projected onto the single-machine safety envelope to obtain a projection point. If the projection point is within the single-machine safety envelope, a control indication is generated based on the control quantity corresponding to the projection point, and this control indication is determined as the target control indication. If the projection point exceeds the single-machine safety envelope, the single-machine candidate control indication is determined as the target control indication. In some examples, if any control quantity in the field-level candidate control indication exceeds the single-machine safety envelope in the safety control auxiliary data, a quadratic programming projection method can be used to determine the target control indication. Alternatively, the control quantities in the field-level candidate control indication can be compared one by one with the corresponding dimensions in the single-machine safety envelope, and the target control indication can be determined based on the control quantities that do not exceed the dimensions and the control quantities that exceed the dimensions. That is, the target control indication is determined through a dimensionally limited degradation scheme.

[0068] Degradation control conditions may include a confidence level in the confidence data representation of the safety control auxiliary data being lower than a confidence threshold. A confidence level lower than the confidence threshold indicates that the confidence level of the field-level candidate control indication corresponding to the confidence data is low. When the confidence level in the confidence data representation is lower than the confidence threshold, the field-level candidate control indication may be rejected. The target control indication determined by the unit edge control device 11 may include a safety limit indication, a single-unit candidate control indication, or a unit default control curve. In one or more embodiments, the confidence threshold may be determined based on the lower quantile of the confidence sample under normal operating conditions. For example, for a normal confidence sample obtained under normal operating conditions according to the same rule, the confidence threshold θC = Q0.01(C normal), where C normal represents a normal confidence sample obtained under normal operating conditions according to the same rule, and Q0.01(C normal) represents the 1st quantile of the normal confidence sample obtained under normal operating conditions according to the same rule.

[0069] For example, Figure 8 A flowchart illustrating an example of determining a target control instruction through degraded control conditions, as provided in this application embodiment, is shown below. Figure 8 As shown, the process of determining the target control instruction through degraded control conditions may include steps S301 to S310.

[0070] In step S301, single-machine candidate control indication, field-level candidate control indication, and safety auxiliary data are acquired.

[0071] In step S302, it is determined whether the security status indicates that a regular security event has been triggered. If a regular security event has been triggered, step S303 is executed. If no regular security event has been triggered, step S304 is executed.

[0072] In step S303, the shutdown instruction or load reduction instruction is determined as the target control instruction.

[0073] In step S304, the communication status indicates whether the communication is abnormal, the channel integrity monitoring result indicates whether the data is diverging, and the indication verification result indicates whether the verification failed. The communication status here indicates the communication between the unit edge control device 11 and the field-level control system 13. If any one of the following conditions is met: communication abnormality, data divergence, or verification failure, then step S305 is executed. If the communication is normal, the data is not diverging, and the verification is successful, then step S306 is executed.

[0074] In step S305, the system switches to single-unit autonomous control. The target control indication for single-unit autonomous control is either a single-unit candidate control indication or the unit's default control curve.

[0075] In step S306, it is determined whether the credibility represented by the credibility data is lower than the credibility threshold. If it is lower than the credibility threshold, step S307 is executed. If it is higher than or equal to the credibility threshold, step S308 is executed.

[0076] In step S307, the safety limit indication, single-unit candidate control indication, or unit default control curve is determined as the target control indication.

[0077] In step S308, the field-level candidate control indication is checked to see if it is within the single-machine safety envelope; if it is, step S309 is executed; if it is not, step S310 is executed.

[0078] In step S309, the field-level candidate control indication is used as the target control indication.

[0079] In step S310, the single-machine safety envelope is projected, and the target control instruction is determined based on the projection.

[0080] The specific details of steps S301 to S310 can be found in the relevant descriptions in the above embodiments, and will not be repeated here. It should be noted that the order of the multiple determinations in steps S301 to S310 can be interchanged, and is not limited here.

[0081] In some embodiments, the unit edge control device 11 can set multiple degradation control conditions and their corresponding control instructions. The degradation control conditions are related to the degree of deterioration of the wind turbine's condition, and different degradation control conditions correspond to different degrees of deterioration of the wind turbine's condition. In this embodiment, if the field-level candidate control instructions are unreasonable, the system can further classify the degradation based on the degree of deterioration of the wind turbine's condition, achieving a stepped degradation control of the wind turbine. The control instruction corresponding to the degradation control conditions satisfied by the safety control auxiliary data or the field-level candidate control instructions can be determined as the target control instruction. For example, the unit edge control device includes four degradation control conditions: L1, L2, L3, and L4 degradation control conditions. The degree of state deterioration of the wind turbines corresponding to L1 to L4 levels gradually increases. The L1 degradation control condition may include an abnormal communication link between the unit edge control device 11 and the field-level control system 13, but the independent control channel 18 between the second control module 112 in the unit edge control device 11 and the wind turbine control device 12 is normal. The control behavior corresponding to the L1 degradation control condition can be to suspend the field-level power generation optimization and execute the candidate indication for safety linkage, that is, to reject the field-level candidate control indication; the L2 degradation control condition... Conditions may include communication status indicating communication anomalies, channel integrity monitoring results indicating data divergence, or indication verification results indicating verification failure. The control behavior corresponding to L2-level degraded control conditions may be executing a single-unit candidate control instruction or the unit's default control curve. L3-level degraded control conditions may include a credibility level represented by credibility data falling below a credibility threshold or a field-level candidate control instruction exceeding the single-unit safety envelope. The control behavior corresponding to L3-level degraded control conditions may be executing a safety limiting instruction or operating within the single-unit safety envelope. L4-level degraded control conditions may include triggering an emergency safety event. When an emergency safety event is triggered, the second control module 112 will take action. The control behavior corresponding to L4-level degraded control conditions may be executing a shutdown instruction or a load reduction instruction. When any level of degraded control condition is triggered, the control instruction corresponding to the satisfied degraded control condition can be determined as the target control instruction and executed. In some examples, the state deterioration of the wind turbine unit gradually deepens. In such cases, different degraded control conditions can be triggered gradually, and the target control instructions corresponding to different degraded control conditions can be executed progressively.

[0082] After executing the target control instruction corresponding to the satisfied degraded control conditions, the wind turbine can be monitored in real time. If the degree of deterioration of the wind turbine's condition decreases, it can be gradually restored to avoid mode oscillations and load shocks caused by repeated switching between different degraded control conditions. Specifically, after executing the target control instruction corresponding to the satisfied degraded control conditions, if the condition monitoring data and operating condition data continuously meet the current recovery control conditions within a preset time period, the control instruction corresponding to the degraded control condition with a less severe degree of deterioration compared to the satisfied degraded control condition is executed. By monitoring the condition monitoring data and operating condition data, the state of the wind turbine after executing the target control instruction can be determined. The recovery control condition can adopt a recovery boundary with an error margin compared to the degraded control condition. Different degraded control conditions have corresponding recovery control conditions. The recovery control condition corresponding to a degraded control condition is the condition that the degree of deterioration of the wind turbine's condition is reduced compared to the degree of deterioration corresponding to this degraded control condition. In addition to conditions related to status monitoring data and operating condition data, the recovery control conditions may also include consistency verification conditions. These consistency verification conditions may include policy version verification of the field-level control system, time synchronization verification, and consistency verification of the wind turbine's state. The preset duration may be the hysteresis duration for jitter prevention. In one or more embodiments, the preset duration may be determined based on the statistical window of relevant data, model observation duration, communication anomaly detection duration, control response stabilization duration, and timing error margin. Specifically, the maximum value among the statistical window of relevant data, model observation duration, communication anomaly detection duration, and control response stabilization duration can be obtained. The maximum value is then summed with the timing error margin and rounded up to an integer multiple of the judgment period to obtain the preset duration. For example, the preset duration may range from 60 seconds to 300 seconds; however, 60 seconds to 300 seconds is merely an example for certain turbine models or operating conditions and is not limited here. If the status monitoring data and operating condition data continuously meet the current recovery control conditions within a preset time period, it indicates that the wind turbine's current state has improved and stabilized. The wind turbine can no longer be controlled according to the control instructions corresponding to the current degraded control conditions. Instead, control instructions corresponding to degraded control conditions with a less severe state deterioration can be used to ensure the wind turbine's stable recovery until it gradually returns to normal operation. This gradual recovery method effectively reduces the number of control mode switching near the threshold, improving the stability of wind turbine operation in the wind farm.

[0083] The control behaviors and recovery behaviors corresponding to different degradation control conditions can be implemented using a state machine. For example, Figure 9 A schematic diagram illustrating an example of a state machine provided in an embodiment of this application, as shown below. Figure 9As shown, the state machine can include five states, L0 to L4. L0 is the normal state, meaning no degradation control condition is triggered. L1 to L4 refer to the degradation control conditions and control behaviors described above. Each state L1 to L4 corresponds to a recovery control condition. When the safety control auxiliary data and the field-level candidate control indication satisfy any degradation control condition in L1 to L4, the system enters the state corresponding to the satisfied degradation control condition and executes the control behavior corresponding to that state. For example, if the unit edge control device 11 detects a heartbeat signal interruption with the field-level control system 13 for more than 10 seconds, or the field-level candidate control indication verification fails, or overlapping data diverges across channels, the system jumps from state L0 to state L2 and executes the single-unit candidate control indication or the unit default control curve corresponding to state L2. As the state deterioration of the wind turbine gradually deepens, the states can also be switched sequentially from L0 to L4. For example, if the safety control auxiliary data and the field-level candidate control indications meet the degradation control conditions corresponding to state L1, then the target control indication corresponding to state L1 is executed. If the state deterioration of the wind turbine continues to worsen, then the system switches to state L2 and executes the target control indication corresponding to state L2. If the state deterioration of the wind turbine continues to worsen, then the system switches to state L3 and executes the target control indication corresponding to state L3. If the state deterioration of the wind turbine continues to worsen, then the system switches to state L4 and executes the target control indication corresponding to state L4. After executing the target control indications, the state deterioration of the wind turbine may be reduced. If the state deterioration of the wind turbine is reduced, the system can switch states step by step in the order from state L4 to state L0. For example, if the safety control auxiliary data and the field-level candidate control indication meet the degradation control conditions corresponding to state L3, then the target control indication corresponding to state L3 is executed. After executing the target control indication corresponding to state L3, if the status monitoring data and operating condition data continuously meet the recovery control conditions corresponding to state L3 within a preset time period, then the system can switch to state L2 and execute the target control indication corresponding to state L2. After executing the target control indication corresponding to state L2, if the status monitoring data and operating condition data continuously meet the recovery control conditions corresponding to state L2 within a preset time period, then the system can switch to state L1 and execute the target control indication corresponding to state L1. After executing the target control indication corresponding to state L1, if the status monitoring data and operating condition data continuously meet the recovery control conditions corresponding to state L1 within a preset time period, then the system can switch to state L0, i.e., restore the normal operation of the wind turbine.

[0084] In some embodiments, the wind farm-level control system 13 can also perform coordinated control of multiple wind turbines in the wind farm. However, it should be noted that this coordinated control also needs to undergo a rationality analysis by the turbine edge control device 11 before determining whether to adopt it. Specifically, the wind farm-level control system 13 can also be used to: upon receiving a safety event sent by the turbine edge control device 11 of any wind turbine in the wind farm, determine the affected turbine set based on the safety event and the correlation between wind turbines in the wind farm. The affected turbine set includes wind turbines affected by the wind turbine in the wind farm where the safety event occurred. The safety event includes regular safety events and emergency safety events. Generate linkage candidate control instructions based on the severity of the safety event and assign linkage candidate control instructions to the turbine edge control devices of the affected turbines in the affected turbine set. Determine the recovery status of the affected wind turbines based on the status monitoring data and operating data of the affected turbines in the affected turbine set. If the recovery status does not meet the predetermined recovery judgment conditions, assign linkage candidate control instructions with higher severity to the turbine edge control devices of the affected turbines in the affected turbine set.

[0085] A safety event is reported by the edge control device 11 of a wind turbine in a wind farm, indicating that a safety event has occurred at that wind turbine. On the one hand, the wind turbine experiencing a safety event may adversely affect other wind turbines in the wind farm; on the other hand, wind turbines in the wind farm with the same or similar characteristics as the wind turbine experiencing the safety event may also experience similar safety events. Correspondingly, after identifying the wind turbine experiencing the safety event, the affected turbine set can be identified promptly, and preventative measures can be taken in advance to improve the safety of the wind turbines in the affected turbine set. The relationships between wind turbines in a wind farm may include, but are not limited to, one or more of the following: turbine model relationship, turbine batch relationship, turbine lifespan relationship, upstream and downstream topology relationship, turbine electrical topology relationship, similar operating conditions relationship, and grid coordination relationship. For example, wind turbines of the same model as those involved in the safety incident in the wind farm can be identified as the affected turbine set; wind turbines from the same batch as those involved in the safety incident can be identified as the affected turbine set; wind turbines with the same service life as those involved in the safety incident can be identified as the affected turbine set; wind turbines with an electrical topology connection to those involved in the safety incident can be identified as the affected turbine set, such as wind turbines connected in series with and sharing the transmission circuit with those involved in the safety incident. The affected wind turbines can be identified as the downstream wind turbines of the wind turbines involved in the safety incident in the wind farm. For example, wind turbines downstream of the wind turbines involved in the safety incident that are affected by the wake can be identified as the affected wind turbines. The affected wind turbines can also be identified as the wind turbines in the wind farm that have similar operating conditions to the wind turbines involved in the safety incident. For example, wind turbines in the wind farm that have similar operating times and conditions to the wind turbines involved in the safety incident can be identified as the affected wind turbines. The affected wind turbines can also be identified as the wind turbines in the wind farm that have grid coordination relationships with the wind turbines involved in the safety incident.

[0086] Candidate control instructions can be generated based on the severity of a safety incident. The higher the severity of the safety incident, the more stringent the generated candidate control instructions. For example, safety incidents can be categorized into six levels of severity, from lowest to highest. The candidate control instruction for level one could be an inspection instruction, level two an instruction to increase monitoring frequency, level three a power limiting instruction, level four a load reduction instruction, level five a yaw correction instruction, and level six a shutdown instruction. It should be noted that the relationship between the severity of a safety incident and the candidate control instructions is not limited to this; the severity levels may be fewer or more. For example, if the temperature and vibration data of the main shaft bearing of a wind turbine in a wind farm simultaneously become abnormal, the second control module 112 can trigger a load reduction instruction based on the wind turbine's preset operating boundary, thus triggering a safety event. After reporting the safety event to the wind farm-level control system 13, the control system 13, based on the correlation between wind turbines in the wind farm, sends preventative inspection instructions to wind turbines in the same batch as the one that triggered the safety event, and sends temporary load reduction instructions to downstream wind turbines of the one that triggered the safety event. Through the safety linkage control of the wind farm, the abnormality of a single wind turbine can be transformed into a wind farm-level preventative control, improving the safety of the wind farm.

[0087] The field-level control system 13 in the above embodiments may further include a field-level linkage module 134. The field-level linkage module 134 can be used to: determine the set of affected units based on the safety event and the correlation between wind turbine units in the wind farm; generate linkage candidate control instructions based on the severity of the safety event; and allocate linkage candidate control instructions to the unit edge control devices of the wind turbine units in the set of affected units.

[0088] For example, Figure 10 A flowchart illustrating an example of field-level safety linkage provided in an embodiment of this application is shown below. Figure 10 As shown, the field-level safety linkage process may include steps a1 to a4.

[0089] In step a1, the sensors in the multi-source sensor group collect abnormal state monitoring data.

[0090] In step a2, abnormal status monitoring data indicates a safety event, which in turn triggers the second control module 112 in the unit edge control device 11 to send a shutdown instruction or load reduction instruction to the wind turbine control device 12.

[0091] In step a3, the wind turbine control device 12 executes a shutdown instruction or a load reduction instruction.

[0092] In step a4, the field-level control system 13 determines the set of affected wind turbines based on safety events and the relationships between wind turbines in the wind farm, and implements coordinated control of the wind turbines in the affected set. Coordinated control may include coordinated control of wind turbines of the same model and batch, coordinated control of wind turbines in upstream and downstream topologies, and coordinated control of wind turbines with grid coordination relationships.

[0093] The specific details of steps a1 to a4 above can be found in the relevant descriptions in the above embodiments, and will not be repeated here.

[0094] The linkage candidate control indication is also a candidate indication, and its rationality needs to be analyzed by the unit edge control device 11 before determining whether to adopt it. In some examples, the unit edge control device 11 can also be used to: determine the target linkage control indication to be allocated to the wind turbine control device 12 based on the single-unit candidate control indication, safety control auxiliary data, and linkage candidate control indication. For the specific content of determining the target linkage control indication based on the single-unit candidate control indication, safety control auxiliary data, and linkage candidate control indication, please refer to the relevant description in the above embodiment on determining the target control indication based on the single-unit candidate control indication, safety control auxiliary data, and farm-level candidate control indication. The linkage candidate control indication can be regarded as the farm-level candidate control indication, and the target linkage control indication can be regarded as the target control indication. The processing flow is basically the same, and will not be repeated here.

[0095] After the field-level control system 13 allocates linkage candidate control instructions to the unit edge control devices 11 affecting the wind turbines in the turbine cluster, and the unit edge control devices 11 determine and execute the target linkage control instruction based on the single-unit candidate control instructions, safety control auxiliary data, and linkage candidate control instructions, the unit edge control devices 11 affecting the wind turbines in the turbine cluster can also provide status monitoring data and operating condition data to the field-level control system 13 to realize the back-collection verification of the target linkage control instruction executed by the field-level control system 13. The field-level control system 13 collects status monitoring data and operating condition data of the wind turbines in the turbine cluster, and determines the recovery status of the wind turbines by analyzing the status monitoring data and operating condition data, that is, whether the recovery status of the wind turbines is a recovery trend. For example, the trend of temperature data, vibration data, load data, wake change data, etc. can be used to determine whether the recovery status of the wind turbines is a recovery trend. The predetermined recovery judgment conditions include the conditions for determining that the recovery status of the wind turbines is a recovery trend. Whether the wind turbine's state is in a recovery trend can be determined by the relationship between the status monitoring data and operating condition data and the predetermined recovery judgment conditions. If the status monitoring data and operating condition data meet the predetermined recovery judgment conditions, the wind turbine's recovery state is in a recovery trend, and the linkage candidate control indication can be maintained. Alternatively, a linkage candidate control indication of lower severity can be assigned to the unit edge control device 11 of the wind turbines in the affected unit set, gradually exiting the degraded control state of this linkage of the wind farm, and gradually restoring the wind turbines in the affected unit set to normal operation. If the status monitoring data and operating condition data do not meet the predetermined recovery judgment conditions, the wind turbine's recovery state is not in a recovery trend, and a linkage candidate control indication of higher severity can be assigned to the unit edge control device 11 of the wind turbines in the affected unit set to increase the control strength of the wind turbines in the affected unit set, and it is expected that the wind turbine's recovery state can be transformed into a recovery trend.

[0096] For example, Figure 11 A flowchart illustrating an example of wind farm safety linkage processing and data recovery verification provided in this application embodiment is shown below. Figure 11 As shown, the process for handling and verifying the safety linkage of the wind farm may include steps S401 to S406.

[0097] In step S401, the set of affected units is determined based on the safety event and the correlation between wind turbine units in the wind farm.

[0098] In step S402, a candidate control instruction for linkage is generated based on the severity of the safety event and sent to the unit edge control device. This candidate control instruction needs to be analyzed for its rationality by the unit edge control device to determine the final target control instruction to be executed.

[0099] In step S403, status monitoring data and operating condition data of the wind turbines in the affected unit set are collected.

[0100] In step S404, it is determined whether the status monitoring data and operating condition data meet the predetermined recovery judgment conditions. First, the recovery status of the wind turbine is determined based on the status monitoring data and operating condition data, and then it is determined whether the recovery status of the wind turbine meets the predetermined recovery judgment conditions. If the predetermined recovery judgment conditions are met, step S405 is executed; if the predetermined recovery judgment conditions are not met, step S406 is executed.

[0101] In step S405, the linkage candidate control indication is maintained, or a recovery candidate control indication is issued. The recovery candidate control indication may be a linkage candidate control indication with lower severity.

[0102] In step S406, the severity of the security event is increased, and step S402 is executed to generate a linkage candidate control indication with a higher severity.

[0103] The specific details of steps S401 to S406 can be found in the relevant descriptions in the above embodiments, and will not be repeated here.

[0104] In some embodiments, the wind farm has a large number of wind turbines, which can be divided into multiple wind turbine clusters, each cluster including at least two wind turbines. Correspondingly, the farm-level control system 13 may include multiple cluster servers and a farm-level server, with the farm-level server communicating with the cluster servers.

[0105] Each cluster server corresponds one-to-one with a wind turbine cluster. The cluster server can generate candidate control instructions based on status monitoring data and operating condition data. It possesses its own cluster control optimization logic, which generates candidate control instructions that achieve the optimization objectives within the cluster control optimization logic, based on the status monitoring data and operating condition data of the wind turbines within the cluster, as well as its own optimization logic. The optimization objectives in the cluster control optimization logic may include, but are not limited to, meeting the demand standard for wind turbine cluster power generation, achieving a balance between wind turbine cluster power generation and wind turbine cluster safety, or meeting the demand standard for wind turbine cluster safety, satisfying grid dispatch constraints, and satisfying load constraints. The candidate control instructions can indicate control parameters, including but not limited to one or more of the following: yaw angle, power allocation value, load reduction ratio, wind turbines shut down, wind turbines started, shutdown time, shutdown batches, start-up sequence, inspection strategy for specific turbines, power limitation for specific turbines, and lockout strategy for specific turbines. The cluster server uploads the generated candidate control instructions to the field-level server.

[0106] The field-level server generates field-level candidate control instructions based on the cluster candidate control instructions generated by each cluster server. The field-level server has its own field-level optimization control strategy, and can generate field-level candidate control instructions that achieve the optimization objectives in the field-level optimization control strategy based on the cluster candidate control instructions generated by each cluster server and the field-level optimization control strategy. The specific content of the field-level control optimization logic can be found in the relevant descriptions in the above embodiments, and will not be repeated here. Cluster candidate control instructions that satisfy the optimization objectives in the cluster optimization control strategy may not necessarily satisfy the optimization objectives in the field-level optimization control strategy. The field-level server can process the cluster candidate control instructions according to the field-level optimization control strategy to obtain the field-level candidate control instructions.

[0107] In some embodiments, such as Figure 2 and Figure 3 As shown, the wind farm's side-field coordinated control system may also include an operation and maintenance (O&M) interactive terminal 17. The O&M interactive terminal 17 is used to interact with users and display information. For example, the O&M interactive terminal 17 may display one or more of the following: wind turbine status monitoring data, the level of degraded control conditions met, the reliability data of each object, safety events that trigger the generation of linkage candidate control instructions, the version of the single-unit control optimization strategy, the trend of wind turbine status changes, and generated alarm information. Figure 6 As shown, the credibility vector can be provided to the operation and maintenance interactive terminal 17, enabling the terminal to display the credibility vector, that is, to display the credibility data in the credibility vector. By displaying information through the operation and maintenance interactive terminal 17, users can gain a more intuitive understanding of the side-field coordinated control of the wind farm, thus improving the user experience.

[0108] This application also provides a side-field coordinated control method for a wind farm, which can be applied to the side-field coordinated control system of the wind farm in the above embodiments. Figure 12 A flowchart of a wind farm sidefield coordinated control method provided in an embodiment of this application is shown below. Figure 12 As shown, the side-field coordinated control method of the wind farm may include steps S501 to S504.

[0109] In step S501, the unit edge control device acquires the status monitoring data and operating condition data of the wind turbine, and transmits the status monitoring data and operating condition data to the field-level control system.

[0110] In step S502, the field-level control system generates field-level candidate control instructions based on the status monitoring data and operating condition data, and sends them to the unit edge control device.

[0111] In step S503, the unit edge control device determines the target control instruction to be allocated to the wind turbine control device based on the individual unit candidate control instruction and the farm-level candidate control instruction. The individual unit candidate control instruction is generated by the unit edge control device based on condition monitoring data and operating condition data.

[0112] In step S504, the wind turbine control device responds to the target control instruction output by the turbine edge control device and drives the wind turbine actuator to perform the action indicated by the target control instruction.

[0113] In some embodiments, the step of the unit edge control device acquiring the status monitoring data and operating condition data of the wind turbine and transmitting the status monitoring data and operating condition data to the farm-level control system can be further refined as follows: the unit edge control device transmits the status monitoring data and operating condition data to the farm-level control system through a first data channel in the data isolation communication device. The step of the farm-level control system issuing commands to the unit edge control device can be further refined as follows: generating a farm-level candidate control instruction and issuing it to the unit edge control device through a second data channel.

[0114] In some embodiments, the unit edge control device includes a first control module and a second control module. The wind farm edge-field coordinated control method may further include: generating a single-unit candidate control instruction by the first control module based on condition monitoring data, operating condition data, and single-unit optimized control objectives.

[0115] The steps for determining the target control instruction based on the single-unit candidate control instruction and the field-level candidate control instruction by the unit edge control device can be further refined as follows: the target control instruction is obtained by the first control module based on the single-unit candidate control instruction and the field-level candidate control instruction.

[0116] The side-field coordinated control method for the wind farm may further include: upon triggering a first predetermined condition, the second control module transmits a control instruction for the first predetermined condition through an independent control channel, so that the wind turbine control device or actuator executes the control instruction. The first predetermined condition is used to determine the occurrence of an emergency safety event.

[0117] In some embodiments, the step of the unit edge control device determining the allocation of target control instructions to the wind turbine control device based on individual unit candidate control instructions and farm-level candidate control instructions can be refined as follows: acquiring safety control auxiliary data, and determining the allocation of target control instructions to the wind turbine control device based on individual unit candidate control instructions, safety control auxiliary data, and farm-level candidate control instructions. The safety control auxiliary data is used to assist in determining the relative importance of individual unit candidate control instructions and farm-level candidate control instructions.

[0118] In some examples, safety control auxiliary data includes factors that can influence the importance of individual unit candidate control indications and farm-level candidate control indications in wind turbine control.

[0119] The specific content of the safety control auxiliary data can be found in the relevant descriptions in the above embodiments, and will not be repeated here.

[0120] In some embodiments, the side-field coordinated control method of the wind farm may further include: in the event of a unit decision triggering event, the first data channel changes the data transmission parameters. The unit decision triggering event indicates that the state of the wind turbine has changed beyond the normal range.

[0121] In some examples, the unit decision-triggered events include one or more of the following: a safety event, a change in operating condition data exceeding a first change threshold, a change in condition monitoring data exceeding a second change threshold, and a change in the credibility represented by credibility data exceeding a third change threshold.

[0122] In some embodiments, the step of determining the target control indication based on the single-unit candidate control indication, safety control auxiliary data, and field-level candidate control indication can be refined as follows: if the safety control auxiliary data and the field-level candidate control indication do not meet the degradation control conditions, the unit edge control device generates the target control indication based on the field-level candidate control indication; if the safety control auxiliary data or the field-level candidate control indication meets the degradation control conditions, the unit edge control device rejects the field-level candidate control indication and determines the target control indication based on the met degradation control conditions.

[0123] In some embodiments, the unit edge control device is configured with multiple degradation control conditions and their corresponding control instructions. The degradation control conditions are related to the degree of state deterioration of the wind turbine, and different degradation control conditions have corresponding recovery control conditions. The wind farm edge-field coordinated control method further includes: after executing the target control instruction corresponding to the satisfied degradation control condition, if the state monitoring data and operating condition data continuously satisfy the current recovery control condition within a preset time period, the unit edge control device executes the control instruction corresponding to a degradation control condition with a less severe state deterioration than the satisfied degradation control condition.

[0124] In some embodiments, if a safety status representation in the safety control auxiliary data triggers a safety event, the unit edge control device rejects the field-level candidate control instruction. If the channel integrity monitoring result in the safety control auxiliary data indicates data divergence between the data channel in the data isolation communication device and the independent safety control channel, the communication status indicates communication interruption, or the indication verification result indicates failure to verify the field-level candidate control indication, then the unit edge control device rejects the field-level candidate control indication. If any control quantity in the field-level candidate control indication exceeds the single-unit safety envelope in the safety control auxiliary data, the unit edge control device determines the target control indication based on the projection of the field-level candidate indication onto the single-unit safety envelope. If the credibility of the credibility data in the safety control auxiliary data is lower than the credibility threshold, the unit edge control device will reject the field-level candidate control instruction.

[0125] In some embodiments, the wind farm edge-field coordinated control method further includes: upon receiving a safety event sent by the edge control device of any wind turbine in the wind farm, the farm-level control system determines the affected turbine set based on the safety event and the correlation between wind turbines in the wind farm, the affected turbine set including the wind turbines affected by the wind turbine in which the safety event occurred; the farm-level control system generates linkage candidate control indications based on the severity of the safety event and assigns linkage candidate control indications to the edge control devices of the wind turbines in the affected turbine set; the farm-level control system collects status monitoring data and operating condition data of the wind turbines in the affected turbine set; if the collected status monitoring data and operating condition data do not meet the predetermined recovery judgment conditions, the farm-level control system assigns linkage candidate control indications with higher severity to the edge control devices of the wind turbines in the affected turbine set.

[0126] In some embodiments, the wind farm edge field coordinated control method further includes: the unit edge control device determining a target linkage control instruction and allocating it to the unit control device based on single unit candidate control instructions, safety control auxiliary data and linkage candidate control instructions.

[0127] In some embodiments, the wind turbines in a wind farm are divided into multiple wind turbine clusters, and the farm-level control system includes multiple cluster servers and a farm-level controller. The step of the farm-level control system generating farm-level candidate control instructions based on status monitoring data and operating condition data can be further refined as follows: the cluster server generates cluster candidate control instructions based on the status monitoring data and operating condition data, and uploads the cluster candidate control instructions to the farm-level server; the farm-level server generates farm-level candidate control instructions based on the cluster candidate control instructions generated by each cluster server.

[0128] It should be noted that the side-field coordinated control method of this wind farm is the same as the side-field coordinated control system of the wind farm mentioned above. All implementation methods in the above method embodiments are applicable to the embodiments of this method and can achieve the same technical effect.

[0129] This application also provides a computer-readable storage medium storing computer program instructions. When these computer program instructions are executed by a processor, they can implement the wind farm side-field coordinated control method described in the above embodiments and achieve the same technical effect. To avoid repetition, further details are omitted here. The aforementioned computer-readable storage medium may include non-transitory computer-readable storage media, such as read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks, etc., and is not limited thereto.

[0130] This application also provides a computer program product, which includes a computer program. When the computer program is executed by a processor, it can implement the wind farm side-field coordinated control method in the above embodiments and achieve the same technical effect. To avoid repetition, it will not be described again here.

[0131] It should be clarified that the various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. For method embodiments, computer-readable storage medium embodiments, and computer program product embodiments, relevant parts can be referred to the description section of the system embodiments. This application is not limited to the specific steps and structures described above and shown in the figures. Those skilled in the art can make various changes, modifications, and additions, or change the order of steps, after understanding the spirit of this application. Furthermore, for the sake of brevity, detailed descriptions of known methods and techniques are omitted here.

[0132] The aspects of this application have been described above with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It should be understood that each block in the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing apparatus to produce a machine such that these instructions, executable via the processor of the computer or other programmable data processing apparatus, enable the implementation of the functions / actions specified in one or more blocks of the flowchart illustrations and / or block diagrams. Such a processor can be, but is not limited to, a general-purpose processor, a special-purpose processor, a special application processor, or a field-programmable logic circuit. It is also understood that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can also be implemented by dedicated hardware performing the specified functions or actions, or can be implemented by a combination of dedicated hardware and computer instructions.

[0133] Those skilled in the art will understand that the above embodiments are exemplary and not restrictive. Different technical features appearing in different embodiments can be combined to achieve beneficial effects. Based on a study of the drawings, specification, and claims, those skilled in the art should be able to understand and implement other variations of the disclosed embodiments. In the claims, the term "comprising" does not exclude other means or steps; the quantifier "a" does not exclude a plurality; the terms "first" and "second" are used to identify names and not to indicate any particular order. No reference numerals in the claims should be construed as limiting the scope of protection. The functionality of multiple parts appearing in the claims can be implemented by a single hardware or software module. The appearance of certain technical features in different dependent claims does not mean that these technical features cannot be combined to achieve beneficial effects.

Claims

1. A side-field coordinated control system for a wind farm, characterized in that, include: The unit-side control device is installed corresponding to the wind turbine unit; The field-level control system is communicatively connected to the unit-side control device. The unit-side control device includes: Wind turbine control device; The unit edge control device is communicatively connected to the wind turbine control device; The wind turbine control device provides the wind turbine's operating condition data to the unit's edge control device; The unit edge control device generates a single-unit candidate control instruction based on the collected wind turbine status monitoring data and the operating condition data, and transmits the status monitoring data and the operating condition data to the field-level control system. The field-level control system generates field-level candidate control instructions based on the status monitoring data and the operating condition data; The unit edge control device determines the target control instruction based on the single-unit candidate control instruction and the farm-level candidate control instruction, and assigns it to the wind turbine control device so that the wind turbine control device executes the target control instruction.

2. The system according to claim 1, characterized in that, Also includes: A data isolation communication device is used to connect the unit edge control device and the field-level control system. The data isolation communication device includes a first data channel and a second data channel. The first data channel is used for the unit edge control device to transmit data to the field-level control system, and the second data channel is used for the field-level control system to transmit data to the unit edge control device.

3. The system according to claim 1, characterized in that, The unit edge control device includes: The first control module is configured to generate a single-machine candidate control instruction based on the status monitoring data, the operating condition data, and the single-machine optimized control target, and to obtain the target control instruction based on the single-machine candidate control instruction and the field-level candidate control instruction. The second control module has an independent control channel with the wind turbine control device, or an independent control channel with the actuator; when a first predetermined condition is triggered, the second control module transmits a control instruction for the first predetermined condition through the independent control channel, the first predetermined condition being used to determine that an emergency safety event has occurred.

4. The system according to claim 1, characterized in that, The unit edge control device is used to acquire safety control auxiliary data, and based on the single-unit candidate control indication, the safety control auxiliary data, and the farm-level candidate control indication, determines the target control indication and assigns it to the wind turbine control device. The safety control auxiliary data is used to help determine the relative importance of the individual candidate control indications and the field-level candidate control indications.

5. The system according to claim 4, characterized in that, The safety control auxiliary data includes factors that can influence the importance of the single-unit candidate control indication and the field-level candidate control indication in wind turbine control.

6. The system according to claim 2, characterized in that, In the event of a unit decision-triggered event, the first data channel changes the data transmission parameters. The unit decision trigger event indicates that the state of the wind turbine unit has changed beyond the normal range.

7. The system according to claim 4, characterized in that, The unit edge control device is used for: If the safety control auxiliary data and the field-level candidate control indication do not meet the degradation control conditions, the target control indication is generated based on the field-level candidate control indication. If the safety control auxiliary data or the field-level candidate control indication meets the degradation control conditions, the field-level candidate control indication is rejected, and the target control indication is determined based on the met degradation control conditions.

8. The system according to claim 1, characterized in that, The unit edge control device is equipped with multiple degradation control conditions and their corresponding control instructions. The degradation control conditions are related to the degree of deterioration of the wind turbine's condition, and different degradation control conditions have corresponding recovery control conditions. The unit edge control device is also used for: After executing the target control instruction corresponding to the satisfied degradation control condition, if the status monitoring data and the operating condition data continue to meet the current recovery control condition within a preset time period, then execute the control instruction corresponding to the degradation control condition with a less severe state deterioration than the satisfied degradation control condition.

9. The system according to claim 1, characterized in that, The field-level control system is also used for: Upon receiving a safety event from the edge control device of any wind turbine in the wind farm, the affected set of turbines is determined based on the safety event and the correlation between wind turbines in the wind farm. Based on the severity of the safety incident, a linkage candidate control instruction is generated, and the linkage candidate control instruction is assigned to the unit edge control device of the wind turbine in the affected unit set. Collect the status monitoring data and operating condition data of the wind turbines in the affected unit set; If the collected status monitoring data and operating condition data do not meet the predetermined recovery judgment conditions, the linkage candidate control indication with higher severity is assigned to the unit edge control device of the wind turbine in the affected unit set.

10. The system according to claim 9, characterized in that, The unit edge control device is also used for: Based on the single-unit candidate control indication, safety control auxiliary data, and the linkage candidate control indication, the target linkage control indication is determined and allocated to the unit control device.

11. The system according to claim 1, characterized in that, The wind turbines in the wind farm are divided into multiple wind turbine clusters, and the farm-level control system includes: Multiple cluster servers, each corresponding to a wind turbine cluster, are used to generate cluster candidate control instructions based on the status monitoring data and the operating condition data. A field-level server, which is communicatively connected to the cluster server, is used to generate the field-level candidate control indication based on the cluster candidate control indication generated by each cluster server.

12. A method for coordinated control of the sidefields of a wind farm, characterized in that, The method, applied to the side-field coordinated control system of a wind farm as described in any one of claims 1 to 11, comprises: The unit-side control device acquires the status monitoring data and operating condition data of the wind turbine, and transmits the status monitoring data and operating condition data to the field-level control system. The field-level control system generates field-level candidate control instructions based on the status monitoring data and the operating condition data, and sends them to the unit-side control device. The unit-side control device determines and executes the target control instruction based on the single-unit candidate control instruction and the field-level candidate control instruction. The single-unit candidate control instruction is generated by the unit-side control device based on the status monitoring data and the operating condition data.

13. A wind turbine generator set, characterized in that, include: An electronic device, the electronic device including a processor and a memory storing computer program instructions; When the processor executes the computer program instructions, it implements the wind farm side-field coordinated control method as described in claim 12.

14. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer program instructions, which, when executed by a processor, implement the side-field coordinated control method for a wind farm as described in claim 12.

15. A computer program product, characterized in that, It includes a computer program that, when executed by a processor, implements the side-field coordinated control method for the wind farm as described in claim 12.

Images