Dedicated programmable controller for wind seeking control

Abstract

The application provides a special programmable controller for wind seeking control and relates to the technical field of controllers, wherein the special programmable controller comprises an input interface, a communication interface, a memory, a central processing unit and an output interface, the cooperation between a rotation angle and a wind direction angle is represented by a differential vector between time sequence vectors of the rotation angle and the wind direction angle, a cooperative input vector is obtained, meanwhile, the influence of the cooperation characteristics between the wind direction angle and the rotation angle on the output characteristics of a wind driven generator is represented by a response estimation between the cooperative input vector and a time sequence vector of output power, so that a logic value of whether the rotation angle of a yaw motor at a current time point should be increased or decreased is determined based on the response estimation, and a rotation angle control instruction is output. The application improves the accuracy of wind seeking control.

Description

Technical Field

[0001] This application relates to the field of controller technology, and in particular to a dedicated programmable controller for wind-seeking control. Background Technology

[0002] A wind turbine is a device that uses wind energy to generate electricity. It consists of a rotating body mounted on the turbine head, blades mounted on the turbine head, and tail fins mounted on the blades. The central axis of the blades is connected to the generator rotor inside the turbine head.

[0003] When operating, the blades rotate under the influence of wind, thus enabling the generator to produce electricity. In existing technology, when the wind direction changes, the blades are generally passively adjusted by a tail fin installed on them. This method cannot allow the blades to actively adjust to face the wind, resulting in unstable power output from the wind turbine. Summary of the Invention

[0004] This application aims to at least partially address one of the technical problems in the related art.

[0005] Therefore, the first objective of this application is to propose a dedicated programmable controller for wind-seeking control, which solves the technical problem that existing methods cannot actively adjust the wind turbine blades to face the wind direction, resulting in unstable power output from the wind turbine. Considering the output performance of the wind turbine and the coordination between the slewing angle and the real-time wind direction, the influence of the coordinated characteristics between the wind direction and the slewing angle on the output characteristics of the wind turbine is represented by the responsiveness estimation of the cooperative input vector of the slewing angle and the wind direction angle relative to the output power input vector. Based on the responsiveness estimation, the slewing angle control command is obtained, thereby improving the accuracy of wind-seeking control.

[0006] To achieve the above objectives, the first aspect of this application proposes a dedicated programmable controller for wind-seeking control, comprising: an input interface for receiving wind direction angles at multiple predetermined time points, including the current time point, and for receiving slewing angles at multiple predetermined time points; a communication interface for receiving the output power of the wind turbine at multiple predetermined time points via communication with the wind turbine; a memory for storing wind direction angle input vectors at multiple predetermined time points, slewing angle input vectors at multiple predetermined time points, and output power input vectors at multiple predetermined time points; a central processing unit for calculating a response feature vector and a response state matrix based on the wind direction angle input vector, the slewing angle input vector, and the output power input vector, and obtaining a regression feature vector based on the response feature vector and the response state matrix, performing logistic regression on the regression feature vector to obtain a logical value indicating whether the slewing angle of the steering motor at the current time point should increase or decrease; and an output interface for outputting slewing angle control commands based on the logical value.

[0007] Optionally, in one embodiment of this application, the central processing unit includes:

[0008] The calculation unit is used to calculate the response feature vector and response state matrix based on the wind direction angle input vector, the gyration angle input vector and the output power input vector, and to obtain the regression feature vector based on the response feature vector and the response state matrix.

[0009] The logistic regression unit is used to perform logistic regression on the regression feature vector to obtain a logical value indicating whether the rotation angle of the directional motor should be increased or decreased at the current time point.

[0010] Optionally, in one embodiment of this application, the computing unit includes:

[0011] The difference unit is used to calculate the positional difference between the wind direction angle input vector and the rotation angle input vector to obtain the cooperative input vector.

[0012] The response subunit is used to calculate the responsiveness estimate of the cooperative input vector with respect to the output power input vector, and obtain the response feature vector;

[0013] The state transition subunit is used to convert the wind direction angle input vector, the slewing angle input vector, and the output power input vector into the first state control vector, the second state control vector, and the third state control vector, respectively, with the characteristic value distributed in (0, 1).

[0014] The response state matrix sub-unit is used to calculate the response state matrix between the first state control vector, the second state control vector, and the third state control vector.

[0015] The regression eigenvector sub-unit is used to multiply the response state matrix by the response eigenvector to obtain the regression eigenvector.

[0016] Optionally, in one embodiment of this application, the differential numerator unit is specifically used for:

[0017] The wind direction angle input vector and the rotation angle input vector are input into multiple parallel subtractors to obtain a co-input vector.

[0018] Optionally, in one embodiment of this application, the response subunit is specifically used for:

[0019] The response feature vector is obtained by inputting the cooperative input vector and the output power input vector into multiple parallel dividers.

[0020] Optionally, in one embodiment of this application, the state transition subunit is specifically used for:

[0021] The wind direction angle input vector, slewing angle input vector, and output power input vector are respectively input into multiple parallel switches based on a predetermined threshold to convert the wind direction angle input vector, slewing angle input vector, and output power input vector into a first state control vector, a second state control vector, and a third state control vector with values ​​distributed in (0,1).

[0022] Optionally, in one embodiment of this application, in a plurality of parallel switches, each switch has the same first control threshold and an independent second control threshold;

[0023] In multiple parallel switches, the threshold of each switch is an independent second control threshold multiplied by the same first control threshold.

[0024] Optionally, in one embodiment of this application, the response state matrix sub-unit is specifically used for:

[0025] The first state control vector and the second state control vector are arrayed using an XOR gate to obtain the same state control matrix;

[0026] The first state control vector and the third state control vector are arrayed using an XOR gate matrix to obtain the first response control matrix;

[0027] The second response control matrix is ​​obtained by passing the second state control vector and the third state control vector through a matrix array of XNOR gates;

[0028] Calculate the union of the first response control matrix and the second response control matrix;

[0029] The response state matrix is ​​obtained by calculating the intersection of the union of the sets and the set of the same state control matrix.

[0030] Optionally, in one embodiment of this application, the logistic regression unit includes:

[0031] The regression value calculation subunit is used to input the regression feature vector into an adder consisting of multiple parallel multipliers connected in series with multiple parallel multipliers to obtain the regression value.

[0032] The logic value calculation subunit is used to input the regression value into a control switch based on a predetermined threshold to obtain a logic value indicating whether the rotation angle of the steering motor should be increased or decreased at the current time point.

[0033] Optionally, in one embodiment of this application, the weight values ​​of the multiple multipliers are equal.

[0034] The dedicated programmable controller for wind-seeking control in this application solves the technical problem that existing methods cannot actively adjust the wind turbine blades to face the wind, resulting in unstable power output from the wind turbine. Considering the output performance of the wind turbine and the coordination between the slewing angle and the real-time wind direction, the responsiveness estimation of the cooperative input vector of the slewing angle and the wind direction angle relative to the output power input vector is used to represent the influence of the cooperative characteristics between the wind direction and the slewing angle on the output characteristics of the wind turbine. Based on the responsiveness estimation, the slewing angle control command is obtained, thereby improving the accuracy of wind-seeking control.

[0035] Additional aspects and advantages of this application will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of this application. Attached Figure Description

[0036] The above and / or additional aspects and advantages of this application will become apparent and readily understood from the following description of the embodiments taken in conjunction with the accompanying drawings, wherein:

[0037] Figure 1 This is a schematic diagram of a scenario for a dedicated programmable controller for wind-finding control provided in Embodiment 1 of this application;

[0038] Figure 2 This is a structural diagram of a dedicated programmable controller for wind-finding control according to an embodiment of this application;

[0039] Figure 3 This is a schematic diagram of the architecture of a dedicated programmable controller for wind-finding control according to an embodiment of this application;

[0040] Figure 4 This is a block diagram of the central processing unit of a dedicated programmable controller for wind-finding control according to an embodiment of this application.

[0041] Figure 5 This is a flowchart illustrating the operation method of a dedicated programmable controller for wind-finding control, as described in an embodiment of this application. Detailed Implementation

[0042] The embodiments of this application are described in detail below. Examples of these embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain this application, and should not be construed as limiting this application.

[0043] In existing methods, when the wind direction changes, the blade orientation is passively adjusted by a tail fin mounted on the blade. This method cannot actively adjust the blade to face the wind, leading to unstable power output from the wind turbine. Therefore, an optimized wind-finding control scheme for wind turbines is needed. Considering the complex and harsh operating conditions of offshore wind turbines, relatively high robustness in control and monitoring is required. Therefore, industrial controllers using Programmable Logic Controllers (PLCs) are employed for wind-finding control in offshore wind turbines, using PLCs to build dedicated edge-side control chips adapted for wind-finding control of offshore wind turbines.

[0044] However, most current systems utilize foreign general-purpose PLC equipment, and most inventions focus on the wiring and control systems between PLC modules and field devices. This fails to achieve controllability based on the data, algorithms, and control processes within the PLC module, and further optimization of these algorithms and control processes. Therefore, to address these technical issues, a dedicated programmable controller for wind-finding control is needed.

[0045] Specifically, in the technical solution of this application, considering that when performing wind-seeking control, i.e., controlling the rotation angle of the directional motor, it is necessary to consider not only the real-time wind direction but also the output performance of the wind turbine. The coordination between the rotation angle and the real-time wind direction is the core element causing changes in the output performance of the wind turbine. Therefore, in the technical solution of this application, the difference vector between the time-series vector of the rotation angle and the time-series vector of the wind direction angle is used to represent the coordination between the two. At the same time, the responsiveness estimation between the coordinated input vector and the time-series vector of the output power is used to represent the influence of the coordination characteristics between wind direction and rotation angle on the output characteristics of the wind turbine.

[0046] In other words, in the technical solution of this application, the wind direction angle, rotation angle, and output power at multiple predetermined time points, including the current time point, are first received through the input interface and / or communication interface of a dedicated programmable controller used for wind-finding control. Optionally, after constructing the wind direction angle, rotation angle, and output power at multiple predetermined time points, including the current time point, as wind speed angle input vector, rotation angle input vector, and output power input vector, respectively, the above input vectors are stored in the memory of the dedicated programmable controller.

[0047] At the central processing unit (CPU) of the dedicated programmable controller (DPPC), the positional difference between the wind direction angle input vector and the yaw angle input vector is first calculated to obtain the cooperative input vector. For example, the wind direction angle input vector and the yaw angle input vector are input into multiple parallel subtractors to subtract them by their respective positions, thus obtaining the cooperative input vector. Next, the responsiveness estimate of the cooperative input vector relative to the output power input vector is calculated to obtain the response feature vector. For example, the cooperative input vector and the output power input vector are input into multiple parallel dividers to divide them by their positions, thus obtaining the response feature vector.

[0048] Furthermore, in the technical solution of this application, it is considered that there will be data deviation at the original data acquisition end. That is, due to the data acquisition capability of the sensor itself or noise interference, there will be errors in the original data acquisition. Moreover, such errors will be amplified with differential calculation and responsiveness estimation calculation. As a result, if logistic regression is directly performed on the response feature vector to obtain the logical value used to represent whether the rotation angle of the steering motor at the current time point should be increased or decreased, the accuracy of responsiveness control will be reduced.

[0049] To address the aforementioned technical problems, this application attempts to use a Markov-like chain-like state transition approach. It obtains a set representation of the state transition response based on the response form of the response state set relative to the control state set, and uses this representation to correct the numerical values ​​of the response feature vectors. This yields a mapping representation of the response values ​​within the state transition space, thereby uncovering the changing relationships of numerical values ​​under the chain-like temporal expression. This allows for further exploration of deeper state transition representations between numerical values ​​beyond numerical information, improving the accuracy of logistic regression.

[0050] Specifically, at the central processing unit (CPU) of the dedicated programmable controller (PLC), the wind direction angle input vector, gyration angle input vector, and output power input vector are first transformed into first-state control vectors, second-state control vectors, and third-state control vectors with characteristic values ​​distributed as (0, 1), respectively. That is, the numerical data in the wind direction angle input vector, gyration angle input vector, and output power input vector are converted into switch-state value vectors through switch-state control lines. It can be understood that converting the wind direction angle, gyration angle, and output power at each predetermined time point into 0,1 distributed logical values ​​reduces the probability of error propagation from the original data source.

[0051] In a specific example, the wind direction angle input vector, the slewing angle input vector, and the output power input vector are respectively input to multiple parallel switches based on a predetermined threshold to convert the wind direction angle input vector, the slewing angle input vector, and the output power input vector into a first state control vector, a second state control vector, and a third state control vector with values ​​distributed as 0 and 1, respectively.

[0052] In one example of this application, each switch has the same control threshold, thus maintaining consistency in the global state metrics across multiple time points: wind direction angle, rotation angle, and output power. In another example, each switch has a separate control threshold, thus maintaining the adaptability of the global state metrics for wind direction angle (rotation angle or output power) at each time point; that is, the control threshold is adaptively determined based on the wind direction angle (rotation angle or output power) at different time points. In yet another example, the confidence level of the state metric is increased for each time point, and the threshold for each switch is a separate control threshold multiplied by the same control threshold. In this way, not only can the consistency of the global state metrics across multiple time points be effectively maintained, but the adaptability of the global state metrics for wind direction angle (rotation angle or output power) at each time point can also be taken into account.

[0053] At the central processing unit (CPU) of the dedicated programmable controller (CPU) used for wind-finding control, the first and second state control vectors are passed through an array of XOR gates to obtain a common state control matrix. This common state control matrix represents the associated set of states where the wind direction angle and glide angle are in the same state. Simultaneously, the first and third state control vectors are passed through an array of XOR gates to obtain a first response control matrix, and the second and third state control vectors are passed through an array of XOR gates to obtain a second response control matrix. That is, the first and second response control matrices represent the positive response sets of output power relative to the wind direction angle and glide angle, respectively. Furthermore, the union of the first and second response control matrices is calculated, and the intersection of this union with the common state control matrix is ​​calculated to obtain the response state matrix.

[0054] Subsequently, at the central processing unit of the dedicated programmable controller used for wind-seeking control, the response state matrix is ​​multiplied by the response feature vector to obtain the regression feature vector. Thus, based on a Markov-like chain-like state transition method, the set representation of the state transition response is obtained based on the response form of the response state set relative to the control state set. This representation is then used to correct the numerical values ​​of the response feature vector, obtaining a mapping representation of the response values ​​in the state transition space. This allows for the discovery of the changing relationships between values ​​under the chain-like temporal expression, further exploring the deeper transition state representations between values ​​beyond the numerical information, thereby improving the accuracy of logistic regression.

[0055] Accordingly, at the central processing unit of the dedicated programmable controller used for wind-seeking control, logistic regression is performed on the regression feature vector to obtain a logical value indicating whether the slewing angle of the steering motor should increase or decrease at the current time point. Furthermore, the dedicated programmable controller can output a slewing angle control command through its output interface based on the logical value indicating whether the slewing angle of the steering motor should increase or decrease at the current time point.

[0056] Based on this, this application provides a dedicated programmable controller for wind-seeking control, comprising: an input interface for receiving wind direction angles at multiple predetermined time points, including the current time point, and receiving slewing angles at multiple predetermined time points; a communication interface for receiving the output power of the wind turbine at multiple predetermined time points through communication with the wind turbine; a memory for storing wind direction angle input vectors at multiple predetermined time points, slewing angle input vectors at multiple predetermined time points, and output power input vectors at multiple predetermined time points; a central processing unit for calculating a response feature vector and a response state matrix based on the wind direction angle input vector, slewing angle input vector, and output power input vector, obtaining a regression feature vector based on the response feature vector and response state matrix, performing logistic regression on the regression feature vector to obtain a logical value indicating whether the slewing angle of the steering motor should increase or decrease at the current time point; and an output interface for outputting slewing angle control commands based on the logical value. This application solves the technical problem that existing methods cannot actively adjust the wind turbine blades to face the wind, resulting in unstable power output from the wind turbine. Considering the output performance of the wind turbine and the coordination between the slewing angle and the real-time wind direction, the application uses the responsiveness estimation of the cooperative input vector of the slewing angle and the wind direction angle relative to the output power input vector to represent the influence of the cooperative characteristics between the wind direction and the slewing angle on the output characteristics of the wind turbine. Based on the responsiveness estimation, the slewing angle control command is obtained, thereby improving the accuracy of wind-seeking control.

[0057] The following describes a dedicated programmable controller for wind-finding control according to an embodiment of this application, with reference to the accompanying drawings.

[0058] Figure 1 This is a schematic diagram of a scenario for a dedicated programmable controller for wind-finding control provided in Embodiment 1 of this application.

[0059] like Figure 1As shown, in the application scenario of a dedicated programmable controller (PLC) for wind-finding control, the wind direction angle at multiple predetermined time points, including the current time point, is first collected from the wind vane Se1. The rotation angle at multiple predetermined time points is collected from the directional motor G1 by the angle sensor Se2, and the output power at multiple predetermined time points is collected from the wind turbine G2 by the power sensor Se3. Then, the wind direction angle, rotation angle, and output power at multiple predetermined time points, including the current time point, are input to the dedicated PLC for wind-finding control via the input interface In and the communication interface Co. The dedicated PLC for wind-finding control processes the wind direction angle, rotation angle, and output power at multiple predetermined time points to obtain a rotation angle control command based on a logic value indicating whether the rotation angle of the directional motor should increase or decrease at the current time point. The rotation angle control command is output via the output interface Out.

[0060] Figure 2 This is a structural diagram of a dedicated programmable controller for wind-finding control according to an embodiment of this application.

[0061] like Figure 2 As shown, the dedicated programmable controller 100 for wind-finding control includes: an input interface 110, a communication interface 120, a memory 130, a central processing unit 140, an output interface 150, and a power supply 160.

[0062] Input interface 110 is used to receive wind direction angles at multiple predetermined time points, including the current time point, and to receive rotation angles at multiple predetermined time points.

[0063] In this embodiment of the application, the input interface 110 is used to receive wind direction angles from the wind vane at multiple predetermined time points, including the current time point, and to receive rotation angles from the steering motor at multiple predetermined time points.

[0064] The communication interface 120 is used to receive the output power of the wind turbine at multiple predetermined time points through communication with the wind turbine.

[0065] The communication interface 120 in this embodiment is used to receive the output power of the wind turbine at multiple predetermined time points via communication with the wind turbine. That is, the output power at multiple predetermined time points is received through the communication interface of a dedicated programmable controller used for wind-seeking control. The output power at the multiple predetermined time points can be acquired by a power sensor.

[0066] The memory 130 is used to store wind direction angle input vectors for multiple predetermined time points, rotation angle input vectors for multiple predetermined time points, and output power input vectors for multiple predetermined time points.

[0067] The memory 130 in this embodiment includes a register for storing wind direction angle input vectors for wind direction angles at multiple predetermined time points, rotation angle input vectors for rotation angles at multiple predetermined time points, and output power input vectors for output power at multiple predetermined time points. That is, after constructing the wind direction angles at multiple predetermined time points (including the current time point), the rotation angles at multiple predetermined time points, and the output power at multiple predetermined time points as wind speed angle input vectors, rotation angle input vectors, and output power input vectors, respectively, these input vectors are stored in the memory of the dedicated programmable controller.

[0068] The central processing unit 140 is used to calculate the response feature vector and the response state matrix based on the wind direction angle input vector, the slewing angle input vector and the output power input vector, and to obtain the regression feature vector based on the response feature vector and the response state matrix. Logistic regression is performed on the regression feature vector to obtain a logical value that indicates whether the slewing angle of the steering motor should be increased or decreased at the current time point.

[0069] The central processing unit 140 of this application embodiment includes a computing unit and a logic regression unit. The central processing unit (CPU) generally consists of a controller, an arithmetic logic unit (ALU), and registers, all integrated into a single chip. The CPU is connected to the memory unit and input / output interface circuits via a data bus, address bus, and control bus. Like a typical computer, the CPU is the core of the PLC, directing the PLC to operate systematically according to the functions assigned by the system program. The user program and data are pre-stored in memory; when the PLC is in running mode, the CPU executes the user program in a cyclic scanning manner.

[0070] The calculation unit in this embodiment is used to calculate the response feature vector and the response state matrix based on the wind direction angle input vector, the slewing angle input vector and the output power input vector, and to obtain the regression feature vector based on the response feature vector and the response state matrix; the logistic regression unit is used to perform logistic regression on the regression feature vector to obtain a logical value that indicates whether the slewing angle of the steering motor should be increased or decreased at the current time point.

[0071] Output interface 150 is used to output rotation angle control commands based on logic values.

[0072] The output interface 150 of this application embodiment is used to output a rotation angle control command based on a logic value indicating whether the rotation angle of the directional motor should be increased or decreased at the current time point, thereby improving the accuracy of wind-seeking control.

[0073] The dedicated programmable controller for wind-seeking control in this application includes an input interface for receiving wind direction angles at multiple predetermined time points, including the current time point, and for receiving slewing angles at multiple predetermined time points; a communication interface for receiving the output power of the wind turbine at multiple predetermined time points through communication with the wind turbine; a memory for storing wind direction angle input vectors at multiple predetermined time points, slewing angle input vectors at multiple predetermined time points, and output power input vectors at multiple predetermined time points; a central processing unit for calculating a response feature vector and a response state matrix based on the wind direction angle input vector, slewing angle input vector, and output power input vector, obtaining a regression feature vector based on the response feature vector and response state matrix, performing logistic regression on the regression feature vector to obtain a logical value indicating whether the slewing angle of the steering motor should increase or decrease at the current time point; and an output interface for outputting slewing angle control commands based on the logical value. Therefore, this method can solve the technical problem that existing methods cannot actively adjust the wind turbine blades to face the wind, resulting in unstable power output from the wind turbine. Considering the output performance of the wind turbine and the coordination between the slewing angle and the real-time wind direction, the method uses the responsiveness estimation of the cooperative input vector of the slewing angle and the wind direction angle relative to the output power input vector to represent the influence of the cooperative characteristics between the wind direction and the slewing angle on the output characteristics of the wind turbine. Based on the responsiveness estimation, the slewing angle control command is obtained, thereby improving the accuracy of wind-seeking control.

[0074] Further, in this embodiment of the application, the central processing unit includes:

[0075] The calculation unit is used to calculate the response feature vector and response state matrix based on the wind direction angle input vector, the gyration angle input vector and the output power input vector, and to obtain the regression feature vector based on the response feature vector and the response state matrix.

[0076] The logistic regression unit is used to perform logistic regression on the regression feature vector to obtain a logical value indicating whether the rotation angle of the directional motor should be increased or decreased at the current time point.

[0077] Further, in this embodiment of the application, the computing unit includes:

[0078] The difference unit is used to calculate the positional difference between the wind direction angle input vector and the rotation angle input vector to obtain the cooperative input vector.

[0079] The response subunit is used to calculate the responsiveness estimate of the cooperative input vector with respect to the output power input vector, and obtain the response feature vector;

[0080] The state transition subunit is used to convert the wind direction angle input vector, the slewing angle input vector, and the output power input vector into the first state control vector, the second state control vector, and the third state control vector, respectively, with the characteristic value distributed in (0, 1).

[0081] The response state matrix sub-unit is used to calculate the response state matrix between the first state control vector, the second state control vector, and the third state control vector.

[0082] The regression eigenvector sub-unit is used to multiply the response state matrix by the response eigenvector to obtain the regression eigenvector.

[0083] The calculation unit in this embodiment is specifically used to calculate the positional difference between the wind direction angle input vector and the gyration angle input vector to obtain a cooperative input vector; calculate the responsiveness estimate of the cooperative input vector relative to the output power input vector to obtain a response feature vector; convert the wind direction angle input vector, the gyration angle input vector, and the output power input vector into a first state control vector, a second state control vector, and a third state control vector with eigenvalues ​​distributed in (0, 1), respectively; calculate the response state matrix between the first state control vector, the second state control vector, and the third state control vector; and multiply the response state matrix by the response feature vector to obtain a regression feature vector.

[0084] Furthermore, in the embodiments of this application, the differential numerator unit is specifically used for:

[0085] The wind direction angle input vector and the rotation angle input vector are input into multiple parallel subtractors to obtain a co-input vector.

[0086] In this embodiment, the wind direction angle input vector and the rotation angle input vector are input into multiple parallel subtractors so that the wind direction angle input vector and the rotation angle input vector are subtracted at the corresponding positions by the multiple parallel subtractors to obtain a cooperative input vector.

[0087] Furthermore, in the embodiments of this application, the response subunit is specifically used for:

[0088] The response feature vector is obtained by inputting the cooperative input vector and the output power input vector into multiple parallel dividers.

[0089] In this embodiment, the cooperative input vector and the output power input vector are input into multiple parallel dividers to perform positional division on the cooperative input vector and the output power input vector through the multiple parallel dividers to obtain the response feature vector.

[0090] Furthermore, in the embodiments of this application, the state transition subunit is specifically used for:

[0091] The wind direction angle input vector, slewing angle input vector, and output power input vector are respectively input into multiple parallel switches based on a predetermined threshold to convert the wind direction angle input vector, slewing angle input vector, and output power input vector into a first state control vector, a second state control vector, and a third state control vector with values ​​distributed in (0,1).

[0092] Considering that there will be data deviation at the raw data acquisition end, that is, due to the sensor's own data acquisition capability or noise interference, there will be errors in the raw data acquisition, and these errors will be amplified with differential calculation and responsiveness estimation calculation. As a result, if logistic regression is directly performed on the response feature vector to obtain the logical value used to represent whether the rotation angle of the steering motor should be increased or decreased at the current time point, the accuracy of responsiveness control will be reduced.

[0093] To address the aforementioned technical problems, this application attempts to use a Markov-like chain-like state transition approach. It obtains a set representation of the state transition response based on the response form of the response state set relative to the control state set, and uses this representation to correct the numerical values ​​of the response feature vectors. This yields a mapping representation of the response values ​​within the state transition space, thereby uncovering the changing relationships of numerical values ​​under the chain-like temporal expression. This allows for further exploration of deeper state transition representations between numerical values ​​beyond numerical information, improving the accuracy of logistic regression.

[0094] Therefore, this application transforms the wind direction angle input vector, gyration angle input vector, and output power input vector into first-state control vectors, second-state control vectors, and third-state control vectors with eigenvalues ​​distributed in (0, 1), respectively. That is, the wind direction angle input vector, gyration angle input vector, and output power input vector are converted into switch-state value vectors by using switch-state control lines. It is understandable that converting the wind direction angle, gyration angle, and output power at each predetermined time point into (0, 1) distributed logical values ​​can reduce the probability of error propagation at the original data end.

[0095] Furthermore, in the embodiments of this application, in the plurality of parallel switches, each switch has the same first control threshold and an independent second control threshold;

[0096] In multiple parallel switches, the threshold of each switch is an independent second control threshold multiplied by the same first control threshold.

[0097] Each switch in this application has the same control threshold, thus maintaining consistency in the global state measurement across multiple time points: wind direction angle, rotation angle, and output power. Simultaneously, each switch in this application has an individual control threshold, which maintains the adaptability of the global state measurement for wind direction angle (rotation angle or output power) at each time point; that is, the control threshold is adaptively determined based on the wind direction angle (rotation angle or output power) at different time points.

[0098] In this embodiment, the confidence level of the state metric is increased for each time point. The threshold of each switch is a separate control threshold multiplied by the same control threshold. In this way, not only can the consistency of the global state metric between wind direction angles (rotation angles or output power) at multiple time points be effectively maintained, but the adaptability of the global state metric of wind direction angles (rotation angles or output power) at each time point can also be taken into account.

[0099] Furthermore, in the embodiments of this application, the response state matrix sub-unit is specifically used for:

[0100] The first state control vector and the second state control vector are arrayed using an XOR gate to obtain the same state control matrix;

[0101] The first state control vector and the third state control vector are arrayed using an XOR gate matrix to obtain the first response control matrix;

[0102] The second response control matrix is ​​obtained by passing the second state control vector and the third state control vector through a matrix array of XNOR gates;

[0103] Calculate the union of the first response control matrix and the second response control matrix;

[0104] The response state matrix is ​​obtained by calculating the intersection of the union of the sets and the set of the same state control matrix.

[0105] In this embodiment, the same state control matrix represents the associated set of wind direction angle and gyration angle in the same state, and the first response control matrix and the second response control matrix represent the positive response set of output power relative to wind direction angle and gyration angle, respectively.

[0106] In this embodiment, the regression feature vector subunit is used to multiply the response state matrix by the response feature vector to obtain the regression feature vector. Thus, based on a Markov-like chain state transition method, the set representation of the state transition response is obtained based on the response form of the response state set relative to the control state set. This representation is then used to correct the numerical values ​​of the response feature vector, obtaining a mapping representation of the response values ​​in the state transition space. This allows for the discovery of the changing relationships between numerical values ​​under the chain-like time-series expression, further exploring the deeper transition state representations between numerical values ​​beyond the numerical information itself, thereby improving the accuracy of logistic regression.

[0107] Furthermore, in this embodiment of the application, the logistic regression unit includes:

[0108] The regression value calculation subunit is used to input the regression feature vector into an adder consisting of multiple parallel multipliers connected in series with multiple parallel multipliers to obtain the regression value.

[0109] The logic value calculation subunit is used to input the regression value into a control switch based on a predetermined threshold to obtain a logic value indicating whether the rotation angle of the steering motor should be increased or decreased at the current time point.

[0110] Furthermore, in the embodiments of this application, the weight values ​​of the multiple multipliers are equal.

[0111] Figure 3 This is a schematic diagram of the architecture of a dedicated programmable controller for wind-finding control according to an embodiment of this application.

[0112] like Figure 3As shown, firstly, the system receives wind direction angles at multiple predetermined time points, including the current time point, from the anemometer via the input interface, and slewing angles at multiple predetermined time points from the directional motor. Then, it receives the output power of the wind turbine at multiple predetermined time points via the communication interface and communication with the wind turbine generator. Next, the system stores the wind direction angle input vector, the slewing angle input vector, and the output power input vector at multiple predetermined time points in the memory. Then, the data is processed by the central processing unit, including: calculating the positional difference between the wind direction angle input vector and the slewing angle input vector to obtain a cooperative input vector; calculating the responsiveness estimate of the cooperative input vector relative to the output power input vector to obtain a response feature vector; converting the wind direction angle input vector, the slewing angle input vector, and the output power input vector into a first-state control vector, a second-state control vector, and a third-state control vector with eigenvalues ​​distributed in (0, 1), respectively; calculating the response state matrix between the first-state control vector, the second-state control vector, and the third-state control vector; multiplying the response state matrix by the response feature vector to obtain a regression feature vector; and performing logistic regression on the regression feature vector to obtain a logical value indicating whether the slewing angle of the steering motor should increase or decrease at the current time point. Furthermore, through the output interface, based on the logical value indicating whether the slewing angle of the steering motor should increase or decrease at the current time point, a slewing angle control command is output.

[0113] Figure 4 This is a block diagram of the central processing unit of a dedicated programmable controller for wind-finding control, as described in an embodiment of this application.

[0114] like Figure 4 As shown, the central processing unit 140 includes: a computing unit 141, configured to calculate the positional difference between the wind direction angle input vector and the slewing angle input vector to obtain a cooperative input vector; calculate the responsiveness estimate of the cooperative input vector relative to the output power input vector to obtain a response feature vector; convert the wind direction angle input vector, the slewing angle input vector, and the output power input vector into a first state control vector, a second state control vector, and a third state control vector with eigenvalues ​​distributed in (0, 1), respectively; calculate the response state matrix between the first state control vector, the second state control vector, and the third state control vector; multiply the response state matrix by the response feature vector to obtain a regression feature vector; and a logistic regression unit 142, configured to perform logistic regression on the regression feature vector to obtain a logical value representing whether the slewing angle of the steering motor should increase or decrease at the current time point.

[0115] This application also proposes a method for operating a dedicated programmable controller for wind-seeking control.

[0116] Figure 5 This is a flowchart illustrating the operation method of a dedicated programmable controller for wind-finding control, as described in an embodiment of this application.

[0117] like Figure 5 As shown, the operation method of the dedicated programmable controller for wind-finding control includes:

[0118] S110 receives wind direction angles at multiple predetermined time points, including the current time point, from the wind vane via the input interface, and receives rotation angles at multiple predetermined time points from the steering motor.

[0119] S120 receives the output power of the wind turbine at multiple predetermined time points via a communication interface and communication with the wind turbine.

[0120] S130, through the memory, stores the wind direction angle input vector of multiple predetermined time points, the rotation angle input vector of multiple predetermined time points, and the output power input vector of multiple predetermined time points;

[0121] S140, data processing is performed by the central processing unit, including: calculating the positional difference between the wind direction angle input vector and the slewing angle input vector to obtain a cooperative input vector through the calculation unit; calculating the responsiveness estimate of the cooperative input vector relative to the output power input vector to obtain a response feature vector; converting the wind direction angle input vector, the slewing angle input vector, and the output power input vector into a first state control vector, a second state control vector, and a third state control vector with eigenvalues ​​distributed in (0, 1), respectively; calculating the response state matrix between the first state control vector, the second state control vector, and the third state control vector; multiplying the response state matrix by the response feature vector to obtain a regression feature vector; and performing logistic regression on the regression feature vector through the logistic regression unit to obtain a logical value representing whether the slewing angle of the steering motor should increase or decrease at the current time point.

[0122] S150 outputs a slewing angle control command via an output interface, based on a logic value indicating whether the slewing angle of the directional motor should increase or decrease at the current time point.

[0123] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., refer to specific features, structures, materials, or characteristics described in connection with that embodiment or example, which are included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0124] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this application, "multiple" means at least two, such as two, three, etc., unless otherwise explicitly specified.

[0125] Any process or method description in the flowchart or otherwise herein can be understood as representing a module, segment, or portion of code comprising one or more executable instructions for implementing custom logic functions or processes, and the scope of the preferred embodiments of this application includes additional implementations in which functions may be performed not in the order shown or discussed, including substantially simultaneously or in reverse order depending on the functions involved, as should be understood by those skilled in the art to which embodiments of this application pertain.

[0126] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-included system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device. More specific examples (a non-exhaustive list) of computer-readable media include: an electrical connection having one or more wires (electronic device), a portable computer disk drive (magnetic device), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Alternatively, the computer-readable medium may be paper or other suitable media on which the program can be printed, since the program can be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in a computer memory.

[0127] It should be understood that various parts of this application can be implemented using hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented using software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0128] Those skilled in the art will understand that all or part of the steps of the methods in the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, the program includes one or a combination of the steps of the method embodiments.

[0129] Furthermore, the functional units in the various embodiments of this application can be integrated into a processing module, or each unit can exist physically separately, or two or more units can be integrated into a module. The integrated module can be implemented in hardware or as a software functional module. If the integrated module is implemented as a software functional module and sold or used as an independent product, it can also be stored in a computer-readable storage medium.

[0130] The storage medium mentioned above can be a read-only memory, a disk, or an optical disk, etc. Although embodiments of this application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting this application. Those skilled in the art can make changes, modifications, substitutions, and variations to the above embodiments within the scope of this application.

Claims

1. A dedicated programmable controller for wind-finding control, characterized in that, include: The input interface is used to receive the wind direction angles at multiple predetermined time points, including the current time point, and to receive the rotation angles at the multiple predetermined time points. A communication interface for receiving the output power of the wind turbine at the plurality of predetermined time points via communication with the wind turbine; The memory is used to store the wind direction angle input vector, the rotation angle input vector, and the output power input vector at the multiple predetermined time points; The central processing unit is used to calculate the response feature vector and the response state matrix based on the wind direction angle input vector, the slewing angle input vector and the output power input vector, and to obtain the regression feature vector based on the response feature vector and the response state matrix, and to perform logistic regression on the regression feature vector to obtain a logical value that indicates whether the slewing angle of the steering motor should be increased or decreased at the current time point. An output interface is used to output a rotation angle control command based on the logic value; The central processing unit includes: The calculation unit is used to calculate the response feature vector and the response state matrix based on the wind direction angle input vector, the gyration angle input vector and the output power input vector, and to obtain the regression feature vector based on the response feature vector and the response state matrix. The logistic regression unit is used to perform logistic regression on the regression feature vector to obtain a logical value that indicates whether the rotation angle of the directional motor should be increased or decreased at the current time point. The computing unit includes: The difference unit is used to calculate the positional difference between the wind direction angle input vector and the rotation angle input vector to obtain the cooperative input vector; A response subunit is used to calculate the responsiveness estimate of the cooperative input vector relative to the output power input vector to obtain a response feature vector; The state transition subunit is used to convert the wind direction angle input vector, the gyration angle input vector and the output power input vector into a first state control vector, a second state control vector and a third state control vector with characteristic values ​​distributed in (0,1), respectively. The response state matrix sub-unit is used to calculate the response state matrix between the first state control vector, the second state control vector, and the third state control vector. The regression feature vector subunit is used to multiply the response state matrix by the response feature vector to obtain the regression feature vector; The differential molecule unit is specifically used for: The wind direction angle input vector and the rotation angle input vector are input into multiple parallel subtractors to obtain the cooperative input vector; The response subunit is specifically used for: The response feature vector is obtained by inputting the cooperative input vector and the output power input vector into multiple parallel dividers. The state transition subunit is specifically used for: The wind direction angle input vector, the gyration angle input vector, and the output power input vector are respectively input into multiple parallel switches based on a predetermined threshold to convert the wind direction angle input vector, the gyration angle input vector, and the output power input vector into the first state control vector, the second state control vector, and the third state control vector, respectively, with values ​​distributed in (0,1). The response state matrix sub-unit is specifically used for: The first state control vector and the second state control vector are arrayed using an XOR gate to obtain the same state control matrix; The first state control vector and the third state control vector are arrayed using an XNOR gate matrix to obtain the first response control matrix; The second state control vector and the third state control vector are arrayed using an XNOR gate matrix to obtain the second response control matrix; Calculate the union of the first response control matrix and the second response control matrix; The response state matrix is ​​obtained by calculating the intersection of the union of the sets and the set of identical state control matrices.

2. The dedicated programmable controller as described in claim 1, characterized in that, In the plurality of parallel switches, each switch has the same first control threshold and an independent second control threshold; In the plurality of parallel switches, the threshold of each switch is the independent second control threshold multiplied by the same first control threshold.

3. The dedicated programmable controller as described in claim 1, characterized in that, The logistic regression unit includes: The regression value calculation subunit is used to input the regression feature vector into multiple parallel multipliers and an adder connected in series with the multiple parallel multipliers to obtain the regression value. The logic value calculation subunit is used to input the regression value into a control switch based on a predetermined threshold to obtain the logic value used to indicate whether the rotation angle of the directional motor should be increased or decreased at the current time point.

4. The dedicated programmable controller as described in claim 3, characterized in that, The weights of the multiple multipliers are equal.

Images