Laser radar wind measurement system and method based on multi-wavelength fusion and intelligent algorithm

Abstract

The invention discloses a laser radar wind measurement system and method based on multi-wavelength fusion and an intelligent algorithm, and relates to the technical field of meteorological monitoring, and the system is characterized in that the system comprises a multi-wavelength laser transmitter which is used for transmitting at least two kinds of lasers with different wavelengths; the high-precision optical receiver is used for receiving the reflected laser signal; the data acquisition module is used for collecting data of the high-precision optical receiver and making preparations for preprocessing the data; the intelligent data processing module is responsible for analyzing frequency components of the laser signals and preprocessing the collected laser signals, including cleaning, normalization and feature extraction; the multi-wavelength laser transmitter, the high-precision optical receiver, the data acquisition module and the intelligent data processing module are connected in sequence. The technical problem to be solved by the invention is to provide a laser radar wind measurement system and method based on multi-wavelength fusion and an intelligent algorithm so as to solve the problems of low precision, low data processing efficiency and the like in related technologies.

Description

Technical Field

[0001] The present invention relates to the field of meteorological monitoring technology, and in particular to a lidar wind measurement system and method based on multi-wavelength fusion and intelligent algorithms. Background Art

[0002] Global climate change and the frequent occurrence of extreme weather events are placing higher demands on meteorological monitoring technology. Wind fields, as a meteorological element, have a significant impact on aviation, agriculture, energy, the environment, and other fields. Traditional wind field measurement methods, while mature and stable, suffer from limitations such as limited measurement range and an inability to obtain three-dimensional wind field information.

[0003] LiDAR technology, with its advantages of high spatiotemporal resolution and non-contact measurement, shows great potential in wind field measurement, enabling precise measurements of wind speed and direction. However, traditional single-wavelength LiDAR systems are susceptible to atmospheric conditions, which can lead to measurement errors. These methods may have limitations when dealing with complex meteorological conditions, such as low accuracy and inefficient data processing. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to provide a lidar wind measurement system and method based on multi-wavelength fusion and intelligent algorithm to solve the problems of low accuracy and low data processing efficiency existing in related technologies.

[0005] The present invention adopts the following technical solutions to achieve the invention objectives:

[0006] The laser radar wind measurement system and method based on multi-wavelength fusion and intelligent algorithm are characterized by including:

[0007] A multi-wavelength laser transmitter for emitting lasers of at least two different wavelengths, adjusting the laser frequency, pulse width, and rate according to different atmospheric conditions, and collimating and focusing the laser beam to ensure efficient energy transmission;

[0008] High-precision optical receiver, used to receive the reflected laser signal;

[0009] A data acquisition module, used to collect data from the high-precision optical receiver and prepare for data preprocessing;

[0010] The intelligent data processing module is responsible for analyzing the frequency components of the laser signal and pre-processing the collected laser signal, including cleaning, normalization, and feature extraction, to facilitate subsequent analysis and processing to ensure data accuracy and usability. It is also responsible for using the trained machine learning model to identify the characteristics of wind speed and direction.

[0011] The multi-wavelength laser transmitter, the high-precision optical receiver, the data acquisition module and the intelligent data processing module are connected in sequence.

[0012] The wind measurement method of the lidar wind measurement system based on multi-wavelength fusion and intelligent algorithm includes the following steps:

[0013] S1: emits laser pulses of different wavelengths;

[0014] S2: Receives the signal reflected from the atmosphere;

[0015] S3: collect signals;

[0016] S4: preprocessing the collected signals;

[0017] S5: Analyze the processed data;

[0018] S6: Correct the measurement results.

[0019] As a further limitation of this technical solution, S1 includes the following steps:

[0020] S101: emitting laser pulses according to a preset schedule and wavelength setting;

[0021] S102: adjusting the frequency and pulse width of laser pulses of each wavelength to adapt to different atmospheric conditions;

[0022] The time intervals and pulse widths of laser emission at different wavelengths need to be precisely controlled using a simple time control algorithm;

[0023] The time control algorithm formula is:

[0024] t n+1 =t n +Δt (1)

[0025] Where: t n is the time of the nth laser emission;

[0026] Δt is the fixed time interval between two transmissions.

[0027] As a further limitation of this technical solution, S2 includes the following steps:

[0028] S201: Capture the laser signal reflected from the atmosphere;

[0029] S202: Convert the received reflected signal into an electrical signal at the same time.

[0030] As a further limitation of this technical solution, S3 includes the following steps:

[0031] S301: Real-time acquisition of captured laser signals reflected from the atmosphere;

[0032] S302: Time-stamp the signal to ensure data synchronization.

[0033] As a further limitation of this technical solution, S4 includes the following steps:

[0034] S401: Processing the collected signal to analyze the frequency components in the signal;

[0035] S402: Denoise, filter, and format the data to remove high-frequency noise.

[0036] Use the Fast Fourier Transform algorithm to analyze the signal frequency components;

[0037] The fast Fourier transform algorithm formula is:

[0038]

[0039] Where: X(k) is the kth frequency domain representation of the signal;

[0040] x(n) is the nth sampling point of the time domain signal;

[0041] N is the number of sampling points;

[0042] i is the imaginary unit;

[0043] k is the frequency index;

[0044] n is the time index.

[0045] As a further limitation of this technical solution, S5 includes the following steps:

[0046] S501: First, use the trained machine learning model support vector machine to identify the characteristics of wind speed and wind direction;

[0047] S502: Analyze the frequency shift of the signal according to the algorithm to calculate the wind speed;

[0048] S503: Finally, estimate the wind direction based on the signal arrival angle and time;

[0049] Use machine algorithms to classify wind speed and direction data;

[0050]

[0051] subject to y i (W·φ(x i )+b)≥1-§ i (4)

[0052] § i ≥0 (5)

[0053] Where: W and b are the parameters of the decision plane;

[0054] φ(x i ) is the mapping function;

[0055] § i is a slack variable;

[0056] C is the regularization parameter;

[0057] ||W|| 2 is the weight vector W = [w1, w2, ..., w d ]’s L2 norm;

[0058] Subject to is a constraint that defines the rules that the model must follow when optimizing the objective function;

[0059] x i is the input feature vector of the i-th sample;

[0060] y i is the true label of the i-th sample.

[0061] As a further limitation of this technical solution, S6 includes the following steps:

[0062] S601: Comparing the output of the machine learning algorithm with a preset model or historical data;

[0063] S602: Identify and correct any systematic errors or deviations;

[0064] S603: Finally, a statistical method is used to optimize the measurement results;

[0065] Use the least squares method to optimize the measurement results;

[0066] The least squares calculation formula is:

[0067]

[0068] in: is the corrected output;

[0069] f(x i ) is the model prediction;

[0070] arg min represents the parameter value that makes the function reach the minimum value.

[0071] Compared with the prior art, the advantages and positive effects of the present invention are:

[0072] The multi-wavelength lidar technology architecture of the present invention includes a multi-wavelength laser transmitter, a high-precision optical receiver, a data acquisition module, a data processing module, a communication module, and a power management module. By emitting lasers of different wavelengths, the impact of atmospheric disturbances is reduced and the accuracy of data is improved. At the same time, the combination of multi-wavelength technology and intelligent algorithms can optimize the processing flow and improve the efficiency and quality of wind field information extraction, thereby overcoming the limitations that may exist when processing complex and changeable meteorological conditions, such as low accuracy and low data processing efficiency, and thus achieving precise control of meteorological monitoring wind speed and direction data, thereby improving accuracy and data processing efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0073] Figure 1 Schematic diagram of the method of the present invention.

[0074] Figure 2 It is a flowchart of the method of the present invention.

[0075] Figure 3 Flowchart of data preprocessing of the present invention.

[0076] Figure 4 Flowchart of data analysis of the present invention.

[0077] Figure 5 Flow chart of the correction measurement results of the present invention.

[0078] Figure 6 Schematic diagram of the system structure of the present invention.

[0079] In the figure: 100, multi-wavelength laser transmitter, 200, high-precision optical receiver, 300, data acquisition module, 400, intelligent data processing module. DETAILED DESCRIPTION

[0080] A specific embodiment of the present invention is described in detail below with reference to the accompanying drawings, but it should be understood that the protection scope of the present invention is not limited by the specific embodiment.

[0081] The present invention comprises:

[0082] The multi-wavelength laser transmitter 100 is used to emit lasers of at least two different wavelengths, adjust the laser frequency, pulse width and frequency according to different atmospheric conditions, and collimate and focus the laser beam to ensure efficient energy transmission; it provides the wind measurement system with a laser signal source that can adapt to different atmospheric environments.

[0083] The high-precision optical receiver 200 is used to receive the reflected laser signal; receive the laser signal reflected in the atmosphere and convert the laser signal into an electrical signal; use a highly sensitive photoelectric detector to detect and receive the weak reflected signal; and be equipped with a filter to reduce the influence of background light and other interference factors.

[0084] The data acquisition module 300 is used to collect data from the high-precision optical receiver 200 to prepare for data preprocessing; receive signals from the optical receiver in real time; perform analog-to-digital conversion to convert the received analog signals into digital signals; store the digital signals in a buffer and perform time stamping to ensure data synchronization.

[0085] The intelligent data processing module 400 is responsible for analyzing the frequency components of the laser signal and pre-processing the collected laser signal, including cleaning, normalization and feature extraction, to facilitate subsequent analysis and processing to ensure the accuracy and availability of the data; it is responsible for using the trained machine learning model to identify the characteristics of wind speed and direction; receiving digital signals from the data acquisition module; executing digital signal processing algorithms, such as fast Fourier transform, to analyze the frequency components of the signal; and applying filtering algorithms to remove high-frequency noise.

[0086] The multi-wavelength laser transmitter 100 , the high-precision optical receiver 200 , the data acquisition module 300 and the intelligent data processing module 400 are connected in sequence.

[0087] The communication module is responsible for data packaging and data exchange with other system components to ensure the security and integrity of data transmission;

[0088] The power management module provides a stable power supply for the entire system, ensuring that the system can continue to operate during power outages.

[0089] The wind measurement method of the lidar wind measurement system based on multi-wavelength fusion and intelligent algorithm includes the following steps:

[0090] S1: Emitting laser pulses of different wavelengths. Using a multi-wavelength laser transmitter, laser pulses of different wavelengths are emitted to provide a signal source under different conditions for wind measurement.

[0091] Said S1 comprises the following steps:

[0092] S101: emitting laser pulses according to a preset schedule and wavelength setting to ensure the orderliness and regularity of laser emission;

[0093] S102: Adjusting the frequency and pulse width of the laser pulses of each wavelength to adapt to different atmospheric conditions. Adjusting the frequency and pulse width of the laser pulses of each wavelength according to different atmospheric conditions (such as temperature, humidity, air pressure, etc.) enables the laser signal to better adapt to the atmospheric environment and improves the accuracy of wind measurement.

[0094] The time intervals and pulse widths of laser emission at different wavelengths need to be precisely controlled using a simple time control algorithm;

[0095] The time control algorithm formula is:

[0096] t n+1 =t n +Δt (1)

[0097] Where: t n is the time of the nth laser emission;

[0098] Δt is the fixed time interval between two transmissions.

[0099] The time interval and pulse width of laser emission at different wavelengths are precisely controlled through a simple time control algorithm to ensure the accuracy and synchronization of laser emission, providing an accurate time basis for subsequent data processing and analysis.

[0100] S2: Receives the signal reflected from the atmosphere, uses a high-precision optical receiver to receive the laser signal reflected from the atmosphere, and obtains the raw data required for wind measurement.

[0101] The S2 comprises the following steps:

[0102] S201: Captures laser signals reflected from the atmosphere. The high-precision optical receiver can effectively capture the laser signals reflected from the atmosphere and obtain the raw data required for wind measurement.

[0103] S202: Convert the received reflected signal into an electrical signal at the same time, and convert the captured laser signal into an electrical signal in real time, so as to facilitate subsequent data collection and processing and realize signal conversion and transmission.

[0104] S3: Collect signals. The data acquisition module collects the captured laser signals in real time and time-marks the signals to ensure data synchronization and provide accurate data for subsequent data processing.

[0105] The S3 includes the following steps:

[0106] S301: Real-time acquisition of laser signals reflected from the atmosphere. The data acquisition module can collect laser signals captured by high-precision optical receivers in real time to ensure the timeliness and integrity of the data.

[0107] S302: Time-stamp the signal to ensure data synchronization. Time-stamp the collected signal and record the acquisition time of the signal to ensure the synchronization of laser signals of different wavelengths and provide an accurate time reference for subsequent data analysis.

[0108] S4: Preprocess the collected signals, perform signal processing on the collected signals, analyze the frequency components in the signals, and perform noise reduction, filtering, formatting and other operations to remove high-frequency noise, improve the quality of the data, and prepare for subsequent analysis and processing.

[0109] As an optional implementation, Figure 3 This is a flowchart of data preprocessing, which preprocesses the collected data, including:

[0110] S401: Processing the collected signal to analyze the frequency components in the signal. Processing the collected laser signal, analyzing the frequency components of the signal through a fast Fourier transform algorithm, extracting useful information from the signal, and providing a basis for measuring wind speed and direction.

[0111] S402: Perform pre-processing operations such as noise reduction, filtering, and formatting on the data to remove high-frequency noise and other interference components in the signal, improve the quality and availability of the data, and provide clean data for subsequent analysis and processing.

[0112] Use the Fast Fourier Transform algorithm to analyze the signal frequency components;

[0113] The fast Fourier transform algorithm formula is:

[0114]

[0115] Where: X(k) is the kth frequency domain representation of the signal;

[0116] x(n) is the nth sampling point of the time domain signal;

[0117] N is the number of sampling points;

[0118] i is the imaginary unit;

[0119] k is the frequency index;

[0120] n is the time index.

[0121] S5: Analyze the processed data and use the trained machine learning model to identify the characteristics of wind speed and direction. Calculate the wind speed by analyzing the frequency shift of the signal. Estimate the wind direction based on the arrival angle and time of the signal to achieve the measurement of wind speed and direction.

[0122] As an optional implementation, Figure 4The flowchart of data analysis is shown in Figure 2. The pre-processed data are analyzed, including:

[0123] S501: First, a trained machine learning model support vector machine (SVM) is used to identify the characteristics of wind speed and wind direction. The pre-trained machine learning model is used to analyze the processed data to identify the characteristics of wind speed and wind direction, thereby achieving preliminary measurement of wind speed and wind direction.

[0124] S502: Analyzing the frequency shift of the signal according to the algorithm to calculate the wind speed. By analyzing the frequency shift of the signal and calculating the wind speed using the relevant algorithm, the accuracy of the wind speed measurement is further improved.

[0125] S503: Finally, the wind direction is estimated based on the arrival angle and time of the signal to achieve wind direction measurement;

[0126] Use machine algorithms to classify wind speed and direction data;

[0127]

[0128] subject to y i (W·φ(x i )+b)≥1-§ i (4)

[0129] § i ≥0 (5)

[0130] Where: W and b are the parameters of the decision plane;

[0131] φ(x i ) is the mapping function;

[0132] § i is a slack variable;

[0133] C is the regularization parameter;

[0134] ||W|| 2 is the weight vector W = [w1, w2, ..., w d ]’s L2 norm;

[0135] Subject to is a constraint that defines the rules that the model must follow when optimizing the objective function;

[0136] x i is the input feature vector of the i-th sample;

[0137] y i is the true label of the i-th sample.

[0138] S6: Correct the measurement results, compare the output of the machine learning algorithm with the preset model or historical data, identify and correct any systematic errors or biases, and use statistical methods (such as least squares method) to optimize the measurement results to improve the accuracy and reliability of wind measurement.

[0139] As an optional implementation, Figure 5 It is a flow chart for correcting measurement results and comparing analysis results with historical data, including:

[0140] S601: Comparing the output of the machine learning algorithm with a preset model or historical data, comparing the wind speed and wind direction measurement results output by the machine learning algorithm with the preset model or historical data to check for systematic errors or deviations;

[0141] S602: Identify and correct any systematic errors or deviations, identify the systematic errors or deviations in the measurement results, and correct them to improve the accuracy of the measurement results;

[0142] S603: Finally, a statistical method is used to optimize the measurement results. A statistical method (such as the least squares method) is used to optimize the corrected measurement results to further improve the accuracy and reliability of the measurement results and ensure that the performance of the wind measurement system reaches the optimal state.

[0143] Use the least squares method to optimize the measurement results;

[0144] The least squares calculation formula is:

[0145]

[0146] in: is the corrected output;

[0147] f(x i ) is the model prediction;

[0148] arg min represents the parameter value that makes the function reach the minimum value.

[0149] The above disclosure is only a specific embodiment of the present invention, but the present invention is not limited thereto. Any changes that can be conceived by those skilled in the art should fall within the scope of protection of the present invention.

Claims

1. The lidar wind measurement system based on multi-wavelength fusion and intelligent algorithm is characterized by: include: A multi-wavelength laser transmitter (100) is used to transmit lasers of at least two different wavelengths, adjust the frequency, pulse width and frequency of the lasers according to different atmospheric conditions, and collimate and focus the laser beam to ensure efficient energy transmission; A high-precision optical receiver (200) for receiving the reflected laser signal; A data acquisition module (300) is used to collect data from the high-precision optical receiver (200) to prepare for data pre-processing; The intelligent data processing module (400) is responsible for analyzing the frequency components of the laser signal and pre-processing the collected laser signal, including cleaning, normalization and feature extraction, to facilitate subsequent analysis and processing to ensure the accuracy and usability of the data; and is responsible for using the trained machine learning model to identify the characteristics of wind speed and direction; The multi-wavelength laser transmitter (100), the high-precision optical receiver (200), the data acquisition module (300), and the intelligent data processing module (400) are connected in sequence.

2. The wind measurement method of the laser radar wind measurement system based on multi-wavelength fusion and intelligent algorithm according to claim 1 is characterized in that: The following steps are involved: S1: emits laser pulses of different wavelengths; S2: Receives the signal reflected from the atmosphere; S3: collect signals; S4: preprocessing the collected signals; S5: Analyze the processed data; S6: Correct the measurement results.

3. The wind measurement method according to claim 2, wherein: Said S1 comprises the following steps: S101: emitting laser pulses according to a preset schedule and wavelength setting; S102: adjusting the frequency and pulse width of laser pulses of each wavelength to adapt to different atmospheric conditions; The time intervals and pulse widths of laser emission at different wavelengths need to be precisely controlled using a simple time control algorithm; The time control algorithm formula is: t n+1 =t n +Δt (1) Where: t n is the time of the nth laser emission; Δt is the fixed time interval between two transmissions.

4. The wind measurement method according to claim 3, wherein: The S2 comprises the following steps: S201: Capture the laser signal reflected from the atmosphere; S202: Convert the received reflected signal into an electrical signal at the same time.

5. The wind measurement method according to claim 4, characterized in that: The S3 includes the following steps: S301: Real-time acquisition of captured laser signals reflected from the atmosphere; S302: Time-stamp the signal to ensure data synchronization.

6. The wind measurement method according to claim 5, characterized in that: The S4 comprises the following steps: S401: Processing the collected signal to analyze the frequency components in the signal; S402: Denoise, filter, and format the data to remove high-frequency noise. Use the Fast Fourier Transform algorithm to analyze the signal frequency components; The fast Fourier transform algorithm formula is: Where: X(k) is the kth frequency domain representation of the signal; x(n) is the nth sampling point of the time domain signal; N is the number of sampling points; i is the imaginary unit; k is the frequency index; n is the time index.

7. The wind measurement method according to claim 6, characterized in that: The S5 comprises the following steps: S501: First, use the trained machine learning model support vector machine to identify the characteristics of wind speed and wind direction; S502: Analyze the frequency shift of the signal according to the algorithm to calculate the wind speed; S503: Finally, estimate the wind direction based on the signal arrival angle and time; Use machine algorithms to classify wind speed and direction data; subject to y i (W·φ(x i )+b)≥1-§ i (4) § i ≥0 (5) Where: W and b are the parameters of the decision plane; φ(x i ) is the mapping function; § i is a slack variable; C is the regularization parameter; ||W|| 2 is the weight vector W = [w1, w2, ..., w d ]’s L2 norm; Subject to is a constraint that defines the rules that the model must follow when optimizing the objective function; x i is the input feature vector of the i-th sample; y i is the true label of the i-th sample.

8. The wind measurement method according to claim 7, characterized in that: The S6 comprises the following steps: S601: Comparing the output of the machine learning algorithm with a preset model or historical data; S602: Identify and correct any systematic errors or deviations; S603: Finally, a statistical method is used to optimize the measurement results; Use the least squares method to optimize the measurement results; The least squares calculation formula is: in: is the corrected output; f(x i ) is the model prediction; arg min represents the parameter value that makes the function reach the minimum value.

Images