Wind speed prediction method and system based on CMAC
By combining multiple detection data of high-definition cameras and ultrasonic anemometers, the CMAC network generates weighted prediction values, solving the problem of single data source and insufficient prediction accuracy in traditional wind speed prediction methods, and improving the accuracy and flexibility of wind speed prediction.
Patent Information
- Application Number
- CN202411340892.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-25
- Publication Date
- 2025-01-07
AI Technical Summary
The traditional wind speed prediction method based on CMAC relies on a single measurement method, resulting in the outliers and noise in the data affecting the learning prediction ability, the generalization ability is insufficient, and the prediction accuracy and practicality need to be improved.
Using multiple wind speed detection devices and sensors, including high-definition cameras and ultrasonic anemometers, the image and sensing data are comprehensively processed through the CMAC network to generate weighted prediction values to improve the accuracy and flexibility of wind speed prediction.
It enhances the ability of variable data generalization of wind speed prediction, improves prediction accuracy and practicality, makes up for the lack of prediction of traditional CMAC models for unlearned wind speed patterns, and achieves the improvement of multiple prediction accuracy.
Smart Images
Figure CN119269836A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of wind speed prediction, and in particular to a CMAC-based wind speed prediction method and system. Background Art
[0002] Wind speed prediction is an important topic in meteorology, renewable energy industry and many other fields. Since wind speed is affected by many factors, including pressure system, terrain characteristics, air pressure, temperature and surface roughness, accurate wind speed prediction is of great significance for optimizing wind farm energy output, aviation traffic safety, marine transportation and urban air quality management. Wind speed prediction is usually divided into short-term prediction and long-term prediction. Short-term wind speed prediction generally refers to the prediction of wind speed changes within a few minutes to a few hours, which is particularly important for the operation and scheduling of wind farms, load adjustment of power systems and ensuring aviation safety. Long-term wind speed prediction focuses on the trend of wind speed changes from a few days to a few weeks or even longer, which has an important impact on the long-term planning of wind farms, shipping route planning and decision-making in agriculture. In terms of wind speed prediction, CMAC has attracted much attention due to its unique advantages. Traditional linear prediction methods often find it difficult to capture these complex changing characteristics, while CMAC can approximate nonlinear functions through its neuro-fuzzy system, thereby better adapting to the variability of wind speed data. CMAC first preprocesses the original wind speed data, including removing noise, filling missing values and possible smoothing, and then inputs the processed data into the CMAC network for training. The CMAC network gradually adjusts the weights to minimize the prediction error through processes such as quantization, address mapping, associative memory and back propagation, which is suitable for online learning and real-time prediction tasks.
[0003] At present, traditional wind speed prediction methods based on CMAC mostly rely on a single measurement method to collect raw data. The outliers and noise in the raw data directly affect the learning and prediction ability of CMAC. Cleaning the raw data takes a lot of time. Due to the variability of meteorology, measurement errors or environmental factors will limit the generalization ability of CMAC. The prediction accuracy and practicality need to be improved. Summary of the invention
[0004] 1. Technical issues to be resolved In view of the shortcomings of the prior art, the present invention provides a wind speed prediction method and system based on CMAC, which has the advantages of flexible response to variable data, strong generalization ability, high multiple prediction accuracy, etc., and solves the problem that the traditional wind speed prediction method based on CMAC has a single data source and the prediction accuracy and practicality need to be improved.
[0005] (II) Technical solution To achieve the above object, the present invention provides the following technical solution: a wind speed prediction method based on CMAC, comprising the following steps: Step 1: The wind speed detection device is fixedly installed in the target area. A detection rod is fixedly installed at the center of the top surface of the wind speed detection device and is perpendicular to the horizon. A high-definition camera is fixedly installed above the wind speed detection device. The data acquisition module uses the high-definition camera to regularly capture the top surface image of the detection rod at a top-down perspective and form an image data set. Step 2: The ultrasonic anemometer is fixedly installed in the target area. The data acquisition module controls the high-definition camera and the ultrasonic anemometer to work synchronously through the network. The data acquisition module obtains the downwind propagation speed and the upwind propagation speed of the ultrasonic wave through the ultrasonic anemometer and forms a sensor data set. Step 3: The wind speed prediction module is provided with CMAC. CMAC takes the connection point between the bottom of the detection rod and the top surface of the wind speed detection device as the origin, establishes a two-dimensional coordinate system parallel to the top surface of the wind speed detection device, and combines the image data set to capture the coordinate position of the top end point of the detection rod in each image in the two-dimensional coordinate system, marked as ; Step 4: The wind speed prediction module detects the coordinates of the top end point of the pole , calculate and generate physical prediction values , predicting future wind speed values based on physical measurement data; Step 5: The wind speed prediction module calculates and generates the sensor prediction value based on the sensor data set , predict future wind speed values based on ultrasonic measurement data; Step 6: CMAC based on physical prediction value and sensor prediction value , comprehensive calculation generates weighted prediction value , output the final wind speed prediction value.
[0006] Preferably, in step one, the top surface of the wind speed detection device and the bottom surface of the high-definition camera are parallel to the horizon, the top of the detection rod is set to an elastic material, and a wind cup is rotatably connected to the surface of the detection rod and is away from the top end point of the detection rod.
[0007] Preferably, in step 1, the expression of the image data set is , to Corresponding to each image of the top surface of the detection rod taken at a top-down angle and the time point of the shooting, Indicates the time point when the high-definition camera captures the image of the top surface of the detection pole, and the numbers are numbered from 1 to Indicates that the top surface image of the detection rod taken at a top-down angle has Each image in the image dataset is arranged in positive order from early to late according to the time point.
[0008] Preferably, in step 2, the expression of the sensor data set is , to Corresponding to the downwind propagation speed and upwind propagation speed obtained by the ultrasonic anemometer each time, It represents the downwind propagation speed obtained by the ultrasonic anemometer sensor. Indicates the headwind propagation speed obtained by the ultrasonic anemometer sensor, with numbers from 1 to It means that the downwind propagation speed and upwind propagation speed obtained by the ultrasonic anemometer sensor are Group.
[0009] Preferably, in step 3, CMAC substitutes all images in the image data set into a two-dimensional coordinate system, calibrates and overlaps the top edge of the wind speed detection device in each image into a single edge, and then captures the coordinate position of the top end point of the detection rod in each image in the two-dimensional coordinate system, marked as .
[0010] Preferably, in step 4, the physical prediction value The calculation formula is as follows:
[0011]
[0012] In the formula, represents the physical prediction value, Indicates the image data set from the 1st to the The coordinates of the top endpoint of the detection rod in the image, Indicates to The horizontal position value of the coordinates of the top end point of the detection rod is Indicates The vertical position value of the coordinates of the top end point of the image detection rod, Indicates that the first The coordinates of the top end point of the pole are detected from the image ( , ) and the Euclidean distance between the coordinates of the bottom end of the detection rod (0,0), which is the physical measurement data. It indicates the fixed coefficient used in the physical prediction value calculation formula to convert the Euclidean distance of coordinate points into wind speed values. Represents the prediction of future wind speed values based on the physical measurement data in the image dataset.
[0013] Preferably, in step 4, the sensor prediction value The calculation formula is as follows: According to the sensor data set, CMAC sets the propagation speed of the ultrasonic wave emitted by the ultrasonic anemometer under windless conditions to , wind speed is , then the ultrasonic wave propagation speed downwind is , the ultrasonic wave propagation speed against the wind is ; According to the sensor data set, CMAC sets the distance between the ultrasonic transmitter and the ultrasonic receiver of the ultrasonic anemometer to , then the propagation time of the ultrasonic wave in the downwind direction is , the propagation time of ultrasonic wave in the upwind direction is The wind speed prediction module calculates the time difference between the downwind propagation time and the upwind propagation time, which is marked as , whose expression is ;
[0014] In the formula, represents the sensor prediction value, It represents the propagation time of ultrasonic wave of ultrasonic anemometer in the downwind direction. It represents the propagation time of ultrasonic wave of ultrasonic anemometer in the upwind direction. It means that the sensor prediction value is calculated based on the Doppler effect during ultrasonic propagation.
[0015] Preferably, in step 5, the weighted prediction value The calculation formula is as follows:
[0016] In the formula, represents the weighted prediction value, Indicates the weighted prediction value calculation formula for the physical prediction value The weight of The weighted prediction value calculation formula for the sensor prediction value The weight of Indicates that the CMAC algorithm model is based on Weight and The weights are calculated comprehensively to get the weighted prediction value, which is the final wind speed prediction value.
[0017] A wind speed prediction system based on CMAC, comprising a high-definition camera and an ultrasonic anemometer, wherein the high-definition camera is used to capture images, and the ultrasonic anemometer is used to detect the downwind propagation speed and the upwind propagation speed of ultrasonic waves. The wind speed prediction system based on CMAC comprises a wind speed detection device, a data acquisition module and a wind speed prediction module; The wind speed detection device is fixedly installed in the target area, a detection rod is fixedly installed at the center of the top surface of the wind speed detection device and is perpendicular to the horizon, a high-definition camera is fixedly installed directly above the wind speed detection device, the top surface of the wind speed detection device and the bottom surface of the high-definition camera are both parallel to the horizon, the top of the detection rod is set to an elastic material, and a wind cup is rotatably connected to the surface of the detection rod and is away from the top end point of the detection rod; The data acquisition module includes a physical measurement unit and a sensor measurement unit. The physical measurement unit is connected to a high-definition camera through a network to collect an image data set. The sensor measurement unit is connected to an ultrasonic anemometer through a network to collect a sensor data set. The data acquisition module transmits the image data set and the sensor data set to the wind speed prediction module through the network. The wind speed prediction module includes an image analysis unit, a physical analysis unit, a sensor analysis unit and a comprehensive prediction unit. The wind speed prediction module is provided with a CMAC. The image analysis unit uses CMAC to establish a two-dimensional coordinate system with the connection point between the bottom end of the detection rod and the top surface of the wind speed detection device as the origin, and then combines the image data set to capture the coordinate position of the top end point of the detection rod in each image in the two-dimensional coordinate system, which is marked as , and transmitted to the physical analysis unit through the network, and the physical analysis unit detects the top end point coordinates of the rod , calculate and generate physical prediction values and transmitted to the comprehensive prediction unit through the network. The sensor analysis unit generates a sensor prediction value based on the sensor data set. , and transmitted to the comprehensive prediction unit through the network, the comprehensive prediction unit according to the physical prediction value and sensor prediction value , comprehensive calculation generates weighted prediction value , and output the final wind speed prediction value.
[0018] Preferably, the data acquisition module is set with a fixed duration, and controls the high-definition camera and the ultrasonic anemometer through a network to work synchronously with the fixed duration as a cycle.
[0019] Compared with the prior art, the present invention provides a wind speed prediction method and system based on CMAC, which has the following beneficial effects: 1. The present invention fixes a detection rod at the center of the top surface of the wind speed detection device and makes it perpendicular to the horizon. A high-definition camera is fixedly installed directly above the wind speed detection device. The top surface of the wind speed detection device and the bottom surface of the high-definition camera are both parallel to the horizon. The top of the detection rod is set to an elastic material. A wind cup is rotatably connected to the surface of the detection rod and is away from the top end point of the detection rod, ensuring that the top of the detection rod can bend with the size of the wind speed. The greater the wind speed, the higher the bending degree of the top of the detection rod, which not only intuitively reflects the wind speed, but also intuitively reflects the specific wind direction. The data acquisition module uses a high-definition camera to regularly shoot the top surface image of the detection rod at a bird's-eye view and compose an image data set. The ultrasonic anemometer is fixedly installed in the target area. The data acquisition module controls the high-definition camera and the ultrasonic anemometer to work synchronously through the network, ensuring the real-time validity of the physical measurement data and the sensor measurement data. The data acquisition module obtains the downwind propagation speed and upwind propagation speed of the ultrasonic wave through the ultrasonic anemometer and forms a sensor data set. The wind speed prediction module is provided with CMAC. CMAC takes the connection point between the bottom end of the detection rod and the top surface of the wind speed detection device as the origin, establishes a two-dimensional coordinate system parallel to the top surface of the wind speed detection device, and combines the image data set to capture the coordinate position of the top end point of the detection rod in each image in the two-dimensional coordinate system, marked as The wind speed prediction module detects the top end coordinates of the pole , calculate and generate physical prediction values , predicting future wind speed values based on physical measurement data, directly makes up for the defect that the traditional CMAC model cannot accurately predict wind speed patterns that have not been learned. The wind speed prediction module calculates and generates sensor prediction values based on the sensor data set , predicts future wind speed values based on ultrasonic measurement data, and has strong generalization ability to flexibly cope with variable data.
[0020] 2. The present invention sets a fixed time length through the data acquisition module, and controls the high-definition camera and ultrasonic anemometer to work synchronously with a fixed time length as a cycle through the network, and cooperates with CMAC according to the physical prediction value and sensor prediction value , comprehensive calculation generates weighted prediction value , output the final wind speed prediction value, and the multiple predictions have high accuracy. BRIEF DESCRIPTION OF THE DRAWINGS
[0021] Figure 1 It is a step diagram of the method of the present invention; Figure 2 It is a schematic diagram of the system flow of the present invention. DETAILED DESCRIPTION
[0022] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0023] Since traditional wind speed prediction methods based on CMAC mostly rely on a single measurement method to collect raw data, the outliers and noise in the raw data directly affect the learning and prediction ability of CMAC. Cleaning the raw data takes a lot of time. Due to the variability of meteorology, measurement errors or environmental factors will limit the generalization ability of CMAC, and the prediction accuracy and practicality need to be improved. Therefore, a wind speed prediction method and system based on CMAC are provided. Please refer to Figure 1-Figure 2 , the method comprises the following steps: Step 1, fix the wind speed detection device in the target area, a detection rod is fixedly installed at the center of the top surface of the wind speed detection device, and it is perpendicular to the horizon, a high-definition camera is fixedly installed directly above the wind speed detection device, the top surface of the wind speed detection device and the bottom surface of the high-definition camera are parallel to the horizon, the top of the detection rod is set to elastic material, to ensure that the top of the detection rod can bend with the size of the wind speed, the greater the wind speed, the higher the bending degree of the top of the detection rod, which not only intuitively reflects the wind speed, but also intuitively reflects the specific wind direction, the surface of the detection rod is rotatably connected with a wind cup, this connection method allows the wind cup to rotate freely around the detection rod under the action of wind force, thereby converting the wind speed into a measurable physical quantity, and away from the top end point of the detection rod, reducing the influence of the wind pressure difference when the wind cup rotates on the top of the detection rod, the data acquisition module uses a high-definition camera to regularly shoot the top surface image of the detection rod at a top-down angle, and composes an image data set; The expression of the image dataset is , to Corresponding to each image of the top surface of the detection rod taken at a top-down angle and the time point of the shooting, Indicates the time point when the high-definition camera captures the image of the top surface of the detection pole, and the numbers are numbered from 1 to Indicates that the top surface image of the detection rod taken at a top-down angle has Each image in the image dataset is arranged in positive order from early to late according to the time point; Step 2: The ultrasonic anemometer is fixedly installed in the target area. The data acquisition module controls the high-definition camera and the ultrasonic anemometer through the network to work synchronously, ensuring the real-time validity of the physical measurement data and the sensor measurement data. The ultrasonic anemometer has a fast response speed, is not affected by changes in wind direction, has no moving parts, and has a low maintenance cost. The data acquisition module obtains the ultrasonic wave's downwind propagation speed and upwind propagation speed through the ultrasonic anemometer and forms a sensor data set. The expression of the sensor data set is: , to Corresponding to the downwind propagation speed and upwind propagation speed obtained by the ultrasonic anemometer each time, It represents the downwind propagation speed obtained by the ultrasonic anemometer sensor. Indicates the headwind propagation speed obtained by the ultrasonic anemometer sensor, with numbers from 1 to It means that the downwind propagation speed and upwind propagation speed obtained by the ultrasonic anemometer sensor are Group; Step 3. The wind speed prediction module is equipped with CMAC, which is a neural network model that simulates the human cerebellum mechanism. It can effectively identify and utilize measurement data from multiple sources, effectively improving the accuracy and efficiency of subsequent wind speed prediction. CMAC uses the connection point between the bottom of the detection rod and the top surface of the wind speed detection device as the origin, and establishes a two-dimensional coordinate system parallel to the top surface of the wind speed detection device. CMAC substitutes all images in the image data set into the two-dimensional coordinate system, calibrates the edge of the top surface of the wind speed detection device in each image and overlaps it into a single edge, and then captures the coordinate position of the top end point of the detection rod in each image in the two-dimensional coordinate system, marked as , which is convenient for the subsequent rapid calculation of the Euclidean distance between the top end point of the detection rod and the bottom end point of the detection rod; Step 4: The wind speed prediction module detects the coordinates of the top end point of the pole , calculate and generate physical prediction values , based on the physical measurement data, the future wind speed value is predicted, and the calculation formula is as follows:
[0024]
[0025] In the formula, represents the physical prediction value, Indicates the image data set from the 1st to the The coordinates of the top endpoint of the detection rod in the image, Indicates to The horizontal position value of the coordinates of the top end point of the detection rod is Indicates The vertical position value of the coordinates of the top end point of the image detection rod, Indicates that the first The coordinates of the top end point of the pole are detected from the image ( , ) and the Euclidean distance between the coordinates of the bottom end of the detection rod (0,0), which is the physical measurement data. It indicates the fixed coefficient used in the physical prediction value calculation formula to convert the Euclidean distance of coordinate points into wind speed values. It indicates that the future wind speed values are predicted based on the physical measurement data in the image dataset, which directly makes up for the defect that the traditional CMAC model cannot accurately predict the wind speed pattern that has not been learned; Step 5: The wind speed prediction module calculates and generates the sensor prediction value based on the sensor data set , based on ultrasonic measurement data to predict future wind speed values, the calculation formula is as follows: According to the sensor data set, CMAC sets the propagation speed of the ultrasonic wave emitted by the ultrasonic anemometer under windless conditions to , wind speed is , then the ultrasonic wave propagation speed downwind is , the ultrasonic wave propagation speed against the wind is ; According to the sensor data set, CMAC sets the distance between the ultrasonic transmitter and the ultrasonic receiver of the ultrasonic anemometer to , then the propagation time of the ultrasonic wave in the downwind direction is , the propagation time of ultrasonic wave in the upwind direction is The wind speed prediction module calculates the time difference between the downwind propagation time and the upwind propagation time, which is marked as , whose expression is ;
[0026] In the formula, represents the sensor prediction value, It represents the propagation time of ultrasonic wave of ultrasonic anemometer in the downwind direction. It represents the propagation time of ultrasonic wave of ultrasonic anemometer in the upwind direction. It means that the sensor prediction value is calculated based on the Doppler effect during ultrasonic propagation, and it has strong generalization ability to flexibly cope with variable data; Step 6: CMAC based on physical prediction value and sensor prediction value , comprehensive calculation generates weighted prediction value , output the final wind speed forecast value, the calculation formula is as follows:
[0027] In the formula, represents the weighted prediction value, Indicates the weighted prediction value calculation formula for the physical prediction value The weight of The weighted prediction value calculation formula for the sensor prediction value The weight of Indicates that the CMAC algorithm model is based on Weight and The weighted prediction value is obtained by comprehensive calculation, which is the final wind speed prediction value, and the multiple predictions have high accuracy.
[0028] The above-mentioned wind speed prediction method based on CMAC is realized by relying on a wind speed prediction system based on CMAC, which includes a wind speed detection device, a data acquisition module and a wind speed prediction module; The wind speed detection device is fixedly installed in the target area. A detection rod is fixedly installed at the center of the top surface of the wind speed detection device and is perpendicular to the horizon. A high-definition camera is fixedly installed directly above the wind speed detection device. The top surface of the wind speed detection device and the bottom surface of the high-definition camera are both parallel to the horizon. The top of the detection rod is set to an elastic material. A wind cup is rotatably connected to the surface of the detection rod and is away from the top end of the detection rod. The data acquisition module is set with a fixed duration, and controls the high-definition camera and the ultrasonic anemometer through the network to work synchronously with a fixed duration as a cycle, and can perform short-term prediction or long-term prediction according to actual prediction needs. The data acquisition module includes a physical measurement unit and a sensor measurement unit. The physical measurement unit is connected to the high-definition camera through the network to collect image data sets, and the sensor measurement unit is connected to the ultrasonic anemometer through the network to collect sensor data sets. The data acquisition module transmits the image data sets and the sensor data sets to the wind speed prediction module through the network; The wind speed prediction module includes an image analysis unit, a physical analysis unit, a sensor analysis unit and a comprehensive prediction unit. The wind speed prediction module is provided with CMAC. The image analysis unit uses CMAC to establish a two-dimensional coordinate system with the connection point between the bottom end of the detection rod and the top surface of the wind speed detection device as the origin. Then, the coordinate position of the top end point of the detection rod in each image in the two-dimensional coordinate system is captured in combination with the image data set, which is marked as , and transmitted to the physical analysis unit through the network. The physical analysis unit detects the coordinates of the top end point of the rod. , calculate and generate physical prediction values The sensor analysis unit calculates and generates sensor prediction values based on the sensor data set. , and transmitted to the comprehensive prediction unit through the network, which can flexibly cope with variable data and has strong generalization ability. The comprehensive prediction unit can predict the value of physical and sensor prediction value , comprehensive calculation generates weighted prediction value , and output the final wind speed prediction value, with high multiple prediction accuracy.
[0029] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.
Claims
1. A wind speed prediction method based on CMAC, characterized in that: The following steps are involved: Step 1: The wind speed detection device is fixedly installed in the target area. A detection rod is fixedly installed at the center of the top surface of the wind speed detection device and is perpendicular to the horizon. A high-definition camera is fixedly installed above the wind speed detection device. The data acquisition module uses the high-definition camera to regularly capture the top surface image of the detection rod at a top-down perspective and form an image data set. Step 2: The ultrasonic anemometer is fixedly installed in the target area. The data acquisition module controls the high-definition camera and the ultrasonic anemometer to work synchronously through the network. The data acquisition module obtains the downwind propagation speed and the upwind propagation speed of the ultrasonic wave through the ultrasonic anemometer and forms a sensor data set. Step 3: The wind speed prediction module is provided with CMAC. CMAC takes the connection point between the bottom of the detection rod and the top surface of the wind speed detection device as the origin, establishes a two-dimensional coordinate system parallel to the top surface of the wind speed detection device, and combines the image data set to capture the coordinate position of the top end point of the detection rod in each image in the two-dimensional coordinate system, marked as ; Step 4: The wind speed prediction module detects the coordinates of the top end point of the pole , calculate and generate physical prediction values , predicting future wind speed values based on physical measurement data; Step 5: The wind speed prediction module calculates and generates the sensor prediction value based on the sensor data set , predict future wind speed values based on ultrasonic measurement data; Step 6: CMAC based on physical prediction value and sensor prediction value , comprehensive calculation generates weighted prediction value , output the final wind speed prediction value.
2. A wind speed prediction method based on CMAC according to claim 1, characterized in that: In step 1, the top surface of the wind speed detection device and the bottom surface of the high-definition camera are parallel to the horizon, the top of the detection rod is set to an elastic material, and a wind cup is rotatably connected to the surface of the detection rod and is away from the top end point of the detection rod.
3. A wind speed prediction method based on CMAC according to claim 2, characterized in that: In step 1, the expression of the image data set is: , to Corresponding to each image of the top surface of the detection rod taken at a top-down angle and the time point of the shooting, Indicates the time point when the high-definition camera captures the image of the top surface of the detection pole, and the numbers are numbered from 1 to Indicates that the top surface image of the detection rod taken at a top-down angle has Each image in the image dataset is arranged in positive order from early to late according to the time point.
4. The CMAC-based wind speed prediction method according to claim 3, characterized in that: In step 2, the expression of the sensor data set is: , to Corresponding to the downwind propagation speed and upwind propagation speed obtained by the ultrasonic anemometer each time, It represents the downwind propagation speed obtained by the ultrasonic anemometer sensor. Indicates the headwind propagation speed obtained by the ultrasonic anemometer sensor, with numbers from 1 to It means that the downwind propagation speed and upwind propagation speed obtained by the ultrasonic anemometer sensor are Group.
5. The wind speed prediction method based on CMAC according to claim 4, characterized in that: In step 3, CMAC substitutes all images in the image data set into a two-dimensional coordinate system, calibrates and overlaps the top edge of the wind speed detection device in each image into a single edge, and then captures the coordinate position of the top end point of the detection rod in each image in the two-dimensional coordinate system, marked as .
6. A wind speed prediction method based on CMAC according to claim 5, characterized in that: In step 4, the physical prediction value The calculation formula is as follows: In the formula, represents the physical prediction value, Indicates the image data set from the 1st to the The coordinates of the top endpoint of the detection rod in the image, Indicates to The horizontal position value of the coordinates of the top end point of the detection rod is Indicates The vertical position value of the coordinates of the top end point of the image detection rod, Indicates that the first The coordinates of the top end point of the pole are detected from the image ( , ) and the Euclidean distance between the coordinates of the bottom end of the detection rod (0,0), which is the physical measurement data. It indicates the fixed coefficient used in the physical prediction value calculation formula to convert the Euclidean distance of coordinate points into wind speed values. Represents the prediction of future wind speed values based on the physical measurement data in the image dataset.
7. A wind speed prediction method based on CMAC according to claim 6, characterized in that: In step 4, the sensor prediction value The calculation formula is as follows: According to the sensor data set, CMAC sets the propagation speed of the ultrasonic wave emitted by the ultrasonic anemometer under windless conditions to , wind speed is , then the ultrasonic wave propagation speed downwind is , the ultrasonic wave propagation speed against the wind is ; According to the sensor data set, CMAC sets the distance between the ultrasonic transmitter and the ultrasonic receiver of the ultrasonic anemometer to , then the propagation time of the ultrasonic wave in the downwind direction is , the propagation time of ultrasonic wave in the upwind direction is The wind speed prediction module calculates the time difference between the downwind propagation time and the upwind propagation time, which is marked as , whose expression is ; In the formula, represents the sensor prediction value, It represents the propagation time of ultrasonic wave of ultrasonic anemometer in the downwind direction. It represents the propagation time of ultrasonic wave of ultrasonic anemometer in the upwind direction. It means that the sensor prediction value is calculated based on the Doppler effect during ultrasonic propagation.
8. A wind speed prediction method based on CMAC according to claim 7, characterized in that: In step 5, the weighted prediction value The calculation formula is as follows: In the formula, represents the weighted prediction value, Indicates the weighted prediction value calculation formula for the physical prediction value The weight of The weighted prediction value calculation formula for the sensor prediction value The weight of Indicates that the CMAC algorithm model is based on Weight and The weights are calculated comprehensively to get the weighted prediction value, which is the final wind speed prediction value.
9. A wind speed prediction system based on CMAC, applied to the wind speed prediction method based on CMAC according to claims 1-8, comprising a high-definition camera and an ultrasonic anemometer, wherein the high-definition camera is used to capture images, and the ultrasonic anemometer is used to detect the downwind propagation speed and the upwind propagation speed of ultrasonic waves, characterized in that: The CMAC-based wind speed prediction system includes a wind speed detection device, a data acquisition module and a wind speed prediction module; The wind speed detection device is fixedly installed in the target area, a detection rod is fixedly installed at the center of the top surface of the wind speed detection device and is perpendicular to the horizon, a high-definition camera is fixedly installed directly above the wind speed detection device, the top surface of the wind speed detection device and the bottom surface of the high-definition camera are both parallel to the horizon, the top of the detection rod is set to an elastic material, and a wind cup is rotatably connected to the surface of the detection rod and is away from the top end point of the detection rod; The data acquisition module includes a physical measurement unit and a sensor measurement unit. The physical measurement unit is connected to a high-definition camera through a network to collect an image data set. The sensor measurement unit is connected to an ultrasonic anemometer through a network to collect a sensor data set. The data acquisition module transmits the image data set and the sensor data set to the wind speed prediction module through the network. The wind speed prediction module includes an image analysis unit, a physical analysis unit, a sensor analysis unit and a comprehensive prediction unit. The wind speed prediction module is provided with a CMAC. The image analysis unit uses CMAC to establish a two-dimensional coordinate system with the connection point between the bottom end of the detection rod and the top surface of the wind speed detection device as the origin, and then combines the image data set to capture the coordinate position of the top end point of the detection rod in each image in the two-dimensional coordinate system, which is marked as , and transmitted to the physical analysis unit through the network, and the physical analysis unit detects the top end point coordinates of the rod , calculate and generate physical prediction values and transmitted to the comprehensive prediction unit through the network. The sensor analysis unit generates a sensor prediction value based on the sensor data set. , and transmitted to the comprehensive prediction unit through the network, the comprehensive prediction unit according to the physical prediction value and sensor prediction value , comprehensive calculation generates weighted prediction value , and output the final wind speed prediction value.
10. A wind speed prediction system based on CMAC according to claim 9, characterized in that: The data acquisition module is set with a fixed duration, and controls the high-definition camera and the ultrasonic anemometer through a network to work synchronously with the fixed duration as a cycle.