A Transformer-based automatic calibration method and system for three-dimensional laser wind radar
Through the automatic calibration method based on Transformer, the deviation probability is predicted using work history data and use information, automatic calibration of laser wind measurement radar is achieved, solving the problem of self-calibration during work in the existing technology, and improving wind speed and direction The accuracy of the measurement.
Patent Information
- Application Number
- CN202411397429.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-10-09
- Publication Date
- 2024-12-06
- Estimated Expiration
- 2044-10-09
AI Technical Summary
The existing laser wind measurement radar calibration methods can only be manually calibrated during installation and maintenance, and cannot be calibrated by itself during work, resulting in the inability to ensure the accuracy of wind speed and wind direction measurement.
Using the automatic calibration method based on Transformer, by obtaining the working history data and use information of the three-dimensional laser wind measurement radar, input a deviation probability prediction model, predict the deviation probability and perform automatic calibration when the deviation is higher than the threshold.
Automatic calibration of three-dimensional laser wind measurement radar is realized, reducing the probability of excessively distorted wind measurement data and improving the accuracy of wind measurement data.
Smart Images

Figure CN118884413B_ABST
Abstract
Description
Technical Field
[0001] The invention belongs to the technical field of laser wind measurement, and in particular relates to an automatic calibration method and system for a three-dimensional laser wind measurement radar based on a Transformer. Background Art
[0002] Laser wind radar is a remote sensing device that uses laser technology to measure wind speed and direction. It measures wind speed based on the Doppler effect by emitting laser beams and receiving light scattered back by aerosol particles in the atmosphere, that is, determining wind speed and direction by analyzing the frequency changes of reflected light.
[0003] In order to ensure the accuracy of wind speed and wind direction measured by the laser wind radar, the laser wind radar needs to be calibrated. The existing calibration methods are all manual calibration during installation and maintenance, and cannot be self-calibrated during the working process, resulting in the accuracy of wind speed and wind direction measured by the laser wind radar cannot be guaranteed. The technical solution designed by the present invention is intended to solve this technical problem. Summary of the invention
[0004] In order to solve the technical problems existing in the above-mentioned background technology, the present invention provides a method, system, electronic equipment, computer storage medium and computer program product for automatic calibration of a three-dimensional laser wind radar based on Transformer.
[0005] The present invention provides a Transformer-based automatic calibration method for a three-dimensional laser wind radar, comprising the following steps:
[0006] Acquire the working history data of the three-dimensional laser wind measurement radar within a preset time period, wherein the working history data includes multiple sets of wind measurement data, and the wind measurement data includes wind speed data and wind direction data;
[0007] Obtaining usage information of the wind measurement data by a user, and determining a deviation tolerance coefficient according to the usage information;
[0008] Inputting the work history data and the deviation tolerance coefficient into a deviation probability prediction model based on Transformer, wherein the deviation probability prediction model predicts a corresponding deviation probability;
[0009] When the deviation probability is higher than the probability threshold, the laser wind measuring radar is controlled to perform automatic calibration in advance.
[0010] In some embodiments, the step of obtaining the operating history data of the three-dimensional laser wind radar within a preset period of time includes:
[0011] Connecting to a third-party meteorological publishing network port to obtain wind forecast information for the area where the three-dimensional laser wind radar is located;
[0012] The terrain data of the area where the three-dimensional laser wind measuring radar is located is obtained based on the high-precision map, and the wind change influence coefficient is evaluated according to the terrain data; wherein the wind change influence coefficient is used to comprehensively characterize the influence of terrain conditions on the speed of change of wind speed and wind direction;
[0013] The duration of the preset time period is determined based on the wind forecast information and the wind change impact coefficient.
[0014] In some embodiments, the determining the duration of the preset time period based on the wind forecast information and the wind change impact coefficient includes:
[0015] Extracting a wind category attribute from the wind forecast information, and determining a wind change assessment value according to the wind category attribute;
[0016] The wind change evaluation value is multiplied by the wind change influence coefficient to obtain a wind change prediction value, and the duration of the preset time period is calculated according to the wind change prediction value and the following calculation formula:
[0017] ;
[0018] In the formula, is the duration of the preset time period, is the baseline value of duration, is the wind change prediction value, is the wind change evaluation value, , is a constant.
[0019] In some embodiments, evaluating and deriving a wind change impact coefficient based on the terrain data includes:
[0020] Calculating the similarity between the terrain data and each preset terrain data, obtaining a preset wind change influence coefficient corresponding to the preset terrain data with the highest similarity, and using it as the wind change influence coefficient of the terrain data;
[0021] The preset wind change influence coefficient corresponding to the preset terrain data is determined in the following manner:
[0022] Selecting several typical preset terrain data, using simulation computational fluid dynamics software to establish a three-dimensional geometric structure of the preset terrain data, and performing numerical simulation based on multiple sets of input wind data and the three-dimensional geometric structure;
[0023] A plurality of measured wind data located in the three-dimensional geometric structure are extracted from the numerical simulation results, and their average value, i.e., the mean of the measured wind data, is calculated. A ratio of the mean of the measured wind data to the corresponding input wind data is calculated, and the preset wind change influence coefficient is determined based on the ratio corresponding to the plurality of the input wind data.
[0024] In some embodiments, the inputting of the work history data and the deviation tolerance coefficient into a deviation probability prediction model based on Transformer, wherein the deviation probability prediction model predicts a corresponding deviation probability, includes:
[0025] Sorting each group of wind measurement data in the work history data in the order of collection time, calculating the difference between adjacent wind measurement data after sorting, and obtaining a wind speed difference matrix and a wind direction difference matrix;
[0026] Input each group of wind measurement data, the wind speed difference matrix, and the wind direction difference matrix into the deviation probability prediction model based on Transformer, the deviation probability prediction model predicts a first deviation probability, and uses the deviation tolerance coefficient to correct the first deviation probability to obtain a second deviation probability;
[0027] The deviation probability prediction model outputs the second deviation probability.
[0028] In some embodiments, determining the deviation tolerance coefficient according to the usage information includes:
[0029] A general large model is used to perform semantic understanding on the usage information, and a data accuracy evaluation is performed on the semantic understanding results based on network search content to determine the deviation tolerance coefficient.
[0030] The present invention also provides a Transformer-based automatic calibration system for a three-dimensional laser wind radar, comprising an acquisition module, a deviation prediction module, and an automatic calibration module;
[0031] The acquisition module is used to acquire the working history data of the three-dimensional laser wind measurement radar within a preset period of time, wherein the working history data includes multiple groups of wind measurement data, wherein the wind measurement data includes wind speed data and wind direction data; and to acquire the usage information of the wind measurement data by the user, and determine the deviation tolerance coefficient according to the usage information;
[0032] The deviation prediction module is used to input the work history data and the deviation tolerance coefficient into a deviation probability prediction model based on Transformer, and the deviation probability prediction model predicts the corresponding deviation probability;
[0033] The automatic calibration module is used to control the laser wind measuring radar to perform automatic calibration in advance when the deviation probability is higher than the probability threshold.
[0034] The present invention also provides an electronic device, comprising: a memory storing executable program code; a processor coupled to the memory; the processor calls the executable program code stored in the memory to execute the method as described in any of the preceding items.
[0035] The present invention also provides a computer storage medium, on which a computer program is stored. When the computer program is executed by a processor, any of the above methods is executed.
[0036] The present invention also provides a computer program product, which includes a computer program stored in a computer storage medium, and when the computer program is executed by a processor of an electronic device, it implements any of the methods described above.
[0037] The solution of the present invention can realize automatic calibration of the three-dimensional laser wind measuring radar, and can accurately predict the expected unexpected deviation of the wind measurement data of the three-dimensional laser wind measuring radar, thereby starting automatic calibration in advance, reducing the probability of measuring excessively distorted wind measurement data, and facilitating the accuracy of the wind measurement data used by the user. BRIEF DESCRIPTION OF THE DRAWINGS
[0038] The accompanying drawings are used to better understand the present solution and do not constitute a limitation of the present disclosure.
[0039] Figure 1 It is a flow chart of an automatic calibration method of a three-dimensional laser wind radar based on Transformer disclosed in an embodiment of the present invention;
[0040] Figure 2 is an explanatory diagram of the changing characteristics of wind speed and wind direction of different types of wind disclosed in an embodiment of the present invention;
[0041] Figure 3 It is a structural schematic diagram of an automatic calibration system of a three-dimensional laser wind radar based on Transformer disclosed in an embodiment of the present invention. DETAILED DESCRIPTION
[0042] The following is a description of exemplary embodiments of the present disclosure in conjunction with the accompanying drawings, including various details of the embodiments of the present disclosure to facilitate understanding, which should be considered as merely exemplary. Therefore, it should be recognized by those of ordinary skill in the art that various changes and modifications may be made to the embodiments described herein without departing from the scope and spirit of the present disclosure. Similarly, for the sake of clarity and conciseness, descriptions of well-known functions and structures are omitted in the following description.
[0043] like Figure 1 As shown, the embodiment of the present invention discloses an automatic calibration method of a three-dimensional laser wind radar based on Transformer, comprising the following steps:
[0044] Acquire the working history data of the three-dimensional laser wind measurement radar within a preset time period, wherein the working history data includes multiple sets of wind measurement data, and the wind measurement data includes wind speed data and wind direction data;
[0045] Obtaining usage information of the wind measurement data by a user, and determining a deviation tolerance coefficient according to the usage information;
[0046] Inputting the work history data and the deviation tolerance coefficient into a deviation probability prediction model based on Transformer, wherein the deviation probability prediction model predicts a corresponding deviation probability;
[0047] When the deviation probability is higher than the probability threshold, the laser wind measuring radar is controlled to perform automatic calibration in advance.
[0048] Compared with the existing calibration method mentioned in the background technology, which can only be manually calibrated during installation and maintenance, the present invention designs a solution that can automatically calibrate during the working process of the laser wind measuring radar. Specifically, the three-dimensional laser wind measuring radar will measure multiple groups of wind measurement data during the working process, each group of wind measurement data contains wind speed data and wind direction data, which constitutes the working history data, and specifically obtains the working history data part located in the preset time period; at the same time, the use information of the user of the wind measurement data for the wind measurement data is also obtained, and the deviation tolerance coefficient of the user for the wind measurement data is determined according to the use information. The deviation tolerance coefficient refers to the degree of deviation between the wind measurement data and the actual wind data that the user can accept. The greater the degree of deviation, the greater the corresponding deviation tolerance coefficient, and vice versa. Finally, the aforementioned acquired working history data and the deviation tolerance coefficient are input into the deviation probability prediction model together, and the deviation probability prediction model predicts the corresponding deviation probability. When the deviation probability is higher than the probability threshold, it is determined that the three-dimensional laser wind measurement radar is about to have an unexpected deviation, and correspondingly, the three-dimensional laser wind measurement radar is controlled to be automatically calibrated in time.
[0049] The solution of the present invention can realize automatic calibration of the three-dimensional laser wind measuring radar, and can accurately predict the expected unexpected deviation of the wind measurement data of the three-dimensional laser wind measuring radar, thereby starting automatic calibration in advance, reducing the probability of measuring excessively distorted wind measurement data, and facilitating the accuracy of the wind measurement data used by the user.
[0050] The deviation probability prediction model used in the present invention is constructed based on the Transformer model. The Transformer model is a deep learning architecture whose core is the self-attention mechanism (Self-Attention), which allows the model to dynamically assign different attention weights between different parts of the sequence, thereby capturing the dependencies within the sequence. This mechanism enables the Transformer to effectively handle long-distance dependency problems and has advantages in parallel computing because it does not need to process the sequence step by step like the traditional recurrent neural network (RNN). In view of the above-mentioned characteristics of the Transformer model, the present invention uses the Transformer model to construct the above-mentioned deviation probability prediction model, which is used to perform in-depth analysis of work history data spanning a long time series, and uses the deviation tolerance coefficient as a reference to predict the deviation probability of the wind measurement data measured by the three-dimensional laser wind measurement radar.
[0051] Among them, the deviation probability predicted by the deviation probability prediction model is a future moment with a specified time length from the current moment. The specified time length can be 2 minutes, 5 minutes, etc., and can be freely set according to the actual situation. And, when it is found that the deviation probability is higher than the probability threshold, the laser wind radar can be controlled to perform automatic calibration. Of course, it can also be automatically calibrated at a suitable moment before the above future moment (for example, when the current wind measurement task has been completed). And, the self-calibration of the three-dimensional laser wind radar in the present invention is mainly hardware calibration, that is, the parameters of key components such as lasers, detectors and turntables are tested and adjusted.
[0052] In some embodiments, the step of obtaining the operating history data of the three-dimensional laser wind radar within a preset period of time includes:
[0053] Connecting to a third-party meteorological publishing network port to obtain wind forecast information for the area where the three-dimensional laser wind radar is located;
[0054] The terrain data of the area where the three-dimensional laser wind measuring radar is located is obtained based on the high-precision map, and the wind change influence coefficient is evaluated according to the terrain data; wherein the wind change influence coefficient is used to comprehensively characterize the influence of terrain conditions on the speed of change of wind speed and wind direction;
[0055] The duration of the preset time period is determined based on the wind forecast information and the wind change impact coefficient.
[0056] In this embodiment, the present invention predicts the probability of unexpected deviations in future time periods based on multiple groups of wind measurement data (i.e., working history data) of the three-dimensional laser wind measurement radar. This requires that the number of groups of wind measurement data should be within a reasonable range, and the determination of this reasonable number of groups is obviously crucial.
[0057] Specifically, the present invention first obtains the wind forecast information of the area where the three-dimensional laser wind radar is located by accessing the third-party meteorological publishing network port. Then, based on the comparison of real-time positioning information and high-precision maps, the terrain data of the area where the three-dimensional laser wind radar is located is determined, and the terrain data is evaluated to obtain the wind change influence coefficient (a value greater than 1). The wind change influence coefficient is used to comprehensively characterize the influence of terrain conditions on the speed of change of wind speed and wind direction. For example, if the terrain is an open plain or sea area, the wind change influence coefficient is low, that is, the change of wind speed and wind direction in the area is less affected by the terrain, and the actual wind speed and wind direction are closer to the wind speed and wind direction in the wind forecast information; and if the terrain is mountainous or there are high buildings, etc., the wind change influence coefficient is high, that is, the change of wind speed and wind direction in the area is greatly affected by the terrain, and the actual wind speed and wind direction are usually far from the wind speed and wind direction in the wind forecast information. Of course, the above is only an example, and is not intended to limit the wind change impact coefficient to be only related to the above two types of terrain. In fact, a variety of terrains with different levels of influence on the changes in wind speed and wind direction can be set as needed, and the details will not be repeated. Finally, the length of the preset time period is determined based on the wind forecast information and wind change impact coefficient determined above. The length of the preset time period refers to the time span from the current moment to the historical time direction. By adjusting the size of the time span, the number of groups of wind measurement data finally obtained can be changed, thereby affecting the subsequent prediction of the probability of deviation.
[0058] The third-party weather release network port mainly refers to the API interface provided by various commercial weather release platforms, such as Moji Weather, Xinzhi Weather, and Hefeng Weather, etc. The present invention does not limit the specific type of the above port.
[0059] In some embodiments, the determining the duration of the preset time period based on the wind forecast information and the wind change impact coefficient includes:
[0060] Extracting a wind category attribute from the wind forecast information, and determining a wind change assessment value according to the wind category attribute;
[0061] The wind change evaluation value is multiplied by the wind change influence coefficient to obtain a wind change prediction value, and the duration of the preset time period is calculated according to the wind change prediction value and the following calculation formula:
[0062] ;
[0063] In the formula, is the duration of the preset time period, is the baseline value of duration, is the wind change prediction value, is the wind change evaluation value, , is a constant.
[0064] In this embodiment, the wind forecast information involves the type of wind in the area where the three-dimensional laser wind radar is located, mainly including steady wind, gust, variable wind, cyclical wind, transient wind, pulsating wind, etc. The speed of change of wind speed and wind direction of these types of wind is different, as shown in the following example. Figure 2 As shown. By conducting targeted evaluation on the above wind category attributes, the wind change evaluation value can be determined, which is used to comprehensively characterize the speed and direction of change of the wind type itself. This evaluation can be achieved through general large models such as ChatGPT, Wenxinyiyan, KIMI, etc.
[0065] After the wind change assessment value is evaluated, it is multiplied by the wind change influence coefficient to obtain the wind change prediction value, and the length of the preset time period is calculated based on the wind change prediction value and the above calculation formula. Referring to the above calculation formula, it can be seen that the wind change prediction value is positively correlated with the length of the preset time period: the larger the wind change prediction value, the faster the wind speed and wind direction in the area where the three-dimensional laser wind radar is located change. This rapid change can easily lead to misjudgment of unexpected deviations, resulting in abnormally high deviation probabilities obtained later. At this time, the longer the duration of the preset time period is set, the more groups of wind measurement data are used to analyze the development trend of the measurement results of the three-dimensional laser wind radar from a longer time span, which is conducive to "suppressing" the abnormally high situation caused by the excessively rapid changes in the wind itself, thereby predicting a more accurate deviation probability.
[0066] When the wind change prediction value is smaller, it means that the wind speed and wind direction in the area where the 3D laser wind radar is located change more slowly. In a slowly changing wind environment, the 3D laser wind radar has a higher probability of measuring abnormal data due to its own inaccuracy rather than wind changes. At this time, the shorter the preset time period is set, the fewer groups of wind measurement data are used to analyze the development trend of the 3D laser wind radar's measurement results, which makes it easier to find abnormal data and thus predict a more accurate deviation probability.
[0067] Therefore, the above scheme of the present invention achieves a balance between the overall analysis angle and the local analysis angle when predicting the deviation probability by adjusting the number of groups of wind measurement data, thereby making the predicted deviation probability more accurate.
[0068] In some embodiments, evaluating and deriving a wind change impact coefficient based on the terrain data includes:
[0069] Calculating the similarity between the terrain data and each preset terrain data, obtaining a preset wind change influence coefficient corresponding to the preset terrain data with the highest similarity, and using it as the wind change influence coefficient of the terrain data;
[0070] The preset wind change influence coefficient corresponding to the preset terrain data is determined in the following manner:
[0071] Selecting several typical preset terrain data, using simulation computational fluid dynamics software to establish a three-dimensional geometric structure of the preset terrain data, and performing numerical simulation based on multiple sets of input wind data and the three-dimensional geometric structure;
[0072] A plurality of measured wind data located in the three-dimensional geometric structure are extracted from the numerical simulation results, and their average value, i.e., the mean of the measured wind data, is calculated. A ratio of the mean of the measured wind data to the corresponding input wind data is calculated, and the preset wind change influence coefficient is determined based on the ratio corresponding to the plurality of the input wind data.
[0073] In this embodiment, a plurality of typical terrains, namely preset terrain data, are predetermined, and corresponding three-dimensional geometric structures are established for these preset terrain data using simulation computational fluid dynamics software (CFD), and multiple sets of input wind data of different wind speeds and wind directions are input into these three-dimensional geometric structures for numerical simulation. Multiple measured wind data located in the three-dimensional geometric structure are then extracted from the numerical simulation results, and the average value of these measured wind data, namely the measured wind data mean, is the "actual" wind data after the input wind enters the three-dimensional geometric structure. The ratio of the multiple measured wind data means to the corresponding input wind data is statistically analyzed (e.g., averaged or weighted averaged), and an equivalent ratio can be obtained, which is used as the preset wind change influence coefficient. The preset wind change influence coefficient characterizes the degree of "amplification" of the input wind by the terrain conditions, that is, the speed of change of wind speed and wind direction is increased.
[0074] By calculating the similarity between the terrain data of the area where the three-dimensional laser wind measurement radar is located and each preset terrain data, the matching preset terrain data can be obtained, and the preset wind change influence coefficient corresponding to the matching preset terrain data is used as the wind change influence coefficient of the terrain data.
[0075] In some embodiments, the inputting of the work history data and the deviation tolerance coefficient into a deviation probability prediction model based on Transformer, wherein the deviation probability prediction model predicts a corresponding deviation probability, includes:
[0076] Sorting each group of wind measurement data in the work history data in the order of collection time, calculating the difference between adjacent wind measurement data after sorting, and obtaining a wind speed difference matrix and a wind direction difference matrix;
[0077] Input each group of wind measurement data, the wind speed difference matrix, and the wind direction difference matrix into the deviation probability prediction model based on Transformer, the deviation probability prediction model predicts a first deviation probability, and uses the deviation tolerance coefficient to correct the first deviation probability to obtain a second deviation probability;
[0078] The deviation probability prediction model outputs the second deviation probability.
[0079] In this embodiment, before inputting the work history data into the deviation probability prediction model, each group of wind measurement data is first sorted in the order of collection time, and the wind speed difference and wind direction difference (vector value) of the sorted adjacent wind measurement data are calculated respectively to obtain the wind speed difference matrix and the wind direction difference matrix. Then, each group of sorted wind measurement data together with the wind speed difference matrix and the wind direction difference matrix are input into the deviation probability prediction model, and the deviation probability prediction model predicts the corresponding first deviation probability. Among them, the above two matrices represent the fluctuation of the wind measurement data, and the deviation probability prediction model can identify the abnormal data and its development trend in the work history data based on this, and then output the corresponding first deviation probability.
[0080] The first deviation probability does not take into account the tolerance of the users of the wind measurement data for distortion. Some users can accept "slight distortion" of the wind measurement data because the accuracy requirements of the wind measurement data in their usage scenarios are not so high. At this time, it is useless to self-calibrate the three-dimensional laser wind measurement radar, especially during the critical period of wind measurement. Self-calibration can easily lead to missed wind measurement windows and data loss. In response to the above problems, the present invention further uses the deviation tolerance coefficient obtained above to correct the first deviation probability, and outputs the corrected second deviation probability for use.
[0081] In some embodiments, determining the deviation tolerance coefficient according to the usage information includes:
[0082] A general large model is used to perform semantic understanding on the usage information, and a data accuracy evaluation is performed on the semantic understanding results based on network search content to determine the deviation tolerance coefficient.
[0083] In this embodiment, the user can send the usage information of the wind measurement data to the three-dimensional laser wind measurement radar or the management server of the three-dimensional laser wind measurement radar in advance. The three-dimensional laser wind measurement radar or its management server calls the general large model to understand the usage information semantically, and searches for relevant network resources based on this, and evaluates the data accuracy of the usage information based on these network resources. The purpose of this evaluation is to analyze the requirements for data accuracy involved in the usage of the wind measurement data by the user. Finally, the corresponding deviation tolerance coefficient is determined based on the requirements for the data accuracy. For example, the deviation tolerance coefficient corresponding to the level of the data accuracy requirement is determined by looking up a table. Obviously, the higher the level of the data accuracy requirement, the lower the corresponding deviation tolerance coefficient, and vice versa.
[0084] like Figure 3 As shown, an embodiment of the present invention discloses an automatic calibration system for a three-dimensional laser wind radar based on a Transformer, including an acquisition module, a deviation prediction module, and an automatic calibration module;
[0085] The acquisition module is used to acquire the working history data of the three-dimensional laser wind measurement radar within a preset period of time, wherein the working history data includes multiple groups of wind measurement data, wherein the wind measurement data includes wind speed data and wind direction data; and to acquire the usage information of the wind measurement data by the user, and determine the deviation tolerance coefficient according to the usage information;
[0086] The deviation prediction module is used to input the work history data and the deviation tolerance coefficient into a deviation probability prediction model based on Transformer, and the deviation probability prediction model predicts the corresponding deviation probability;
[0087] The automatic calibration module is used to control the laser wind measuring radar to perform automatic calibration in advance when the deviation probability is higher than the probability threshold.
[0088] An embodiment of the present invention further discloses an electronic device, comprising: a memory storing executable program code; a processor coupled to the memory; the processor calls the executable program code stored in the memory to execute the method as described in any of the preceding items.
[0089] The embodiment of the present invention further discloses a computer storage medium, on which a computer program is stored. When the computer program is executed by a processor, the method described in any of the above items is executed.
[0090] An embodiment of the present invention further discloses a computer program product, which includes a computer program stored in a computer storage medium, and when the computer program is executed by a processor of an electronic device, any of the above methods is implemented.
[0091] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.
[0092] The program code for implementing the method of the present disclosure may be written in any combination of one or more programming languages. These program codes may be provided to a processor or controller of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that the program code, when executed by the processor or controller, enables the functions / operations specified in the flow chart and / or block diagram to be implemented. The program code may be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.
[0093] In the context of the present disclosure, a machine-readable medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, device, or equipment. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or equipment, or any suitable combination of the foregoing. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.
[0094] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user can provide input to the computer. Other types of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, voice input, or tactile input).
[0095] The systems and techniques described herein may be implemented in a computing system that includes back-end components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes front-end components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such back-end components, middleware components, or front-end components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), and the Internet.
[0096] A computer system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The relationship of client and server is generated by computer programs running on respective computers and having a client-server relationship with each other. The server may be a cloud server, a server of a distributed system, or a server combined with a blockchain.
[0097] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps recorded in this disclosure can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solution of this disclosure can be achieved, and this document is not limited here.
[0098] The above specific implementations do not constitute a limitation on the protection scope of the present disclosure. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present disclosure shall be included in the protection scope of the present disclosure.
Claims
1. An automatic calibration method for 3D laser wind radar based on Transformer. It is characterized in that The steps include: Acquire the working history data of the three-dimensional laser wind measuring radar within a preset time period, wherein the working history data includes multiple sets of wind measurement data, and the wind measurement data includes wind speed data and wind direction data; Obtaining usage information of the wind measurement data by the user, and determining a deviation tolerance coefficient according to the usage information; wherein the deviation tolerance coefficient is the degree of deviation between the wind measurement data and the actual wind data that the user can accept, and the greater the degree of deviation, the greater the corresponding deviation tolerance coefficient, and vice versa; Inputting the work history data and the deviation tolerance coefficient into a deviation probability prediction model based on Transformer, wherein the deviation probability prediction model predicts a corresponding deviation probability; When the deviation probability is higher than the probability threshold, the laser wind measuring radar is controlled to perform automatic calibration in advance.
2. According to the automatic calibration method of a three-dimensional laser wind radar based on Transformer in claim 1, Features: Obtain the working history data of the 3D laser wind radar within a preset period of time, including: Connecting to a third-party meteorological publishing network port to obtain wind forecast information for the area where the three-dimensional laser wind radar is located; The terrain data of the area where the three-dimensional laser wind measuring radar is located is obtained based on the high-precision map, and the wind change influence coefficient is evaluated according to the terrain data; wherein the wind change influence coefficient is used to comprehensively characterize the influence of terrain conditions on the speed of change of wind speed and wind direction; The duration of the preset time period is determined based on the wind forecast information and the wind change impact coefficient.
3. The automatic calibration method of a three-dimensional laser wind radar based on Transformer according to claim 2, Features: Determining the duration of the preset time period based on the wind forecast information and the wind change impact coefficient includes: Extracting a wind category attribute from the wind forecast information, and determining a wind change assessment value according to the wind category attribute; The wind change evaluation value is multiplied by the wind change influence coefficient to obtain a wind change prediction value, and the duration of the preset time period is calculated according to the wind change prediction value and the following calculation formula: Wherein, Δt is the duration of the preset time period, T is the reference value of the duration, w2 is the wind change prediction value, w1 is the wind change evaluation value, and n1 and n2 are constants.
4. The automatic calibration method of a three-dimensional laser wind radar based on Transformer according to claim 3, Features: The wind change impact coefficients are evaluated based on the terrain data, including: Calculating the similarity between the terrain data and each preset terrain data, obtaining a preset wind change influence coefficient corresponding to the preset terrain data with the highest similarity, and using it as the wind change influence coefficient of the terrain data; The preset wind change influence coefficient corresponding to the preset terrain data is determined in the following manner: Selecting several typical preset terrain data, using simulation computational fluid dynamics software to establish a three-dimensional geometric structure of the preset terrain data, and performing numerical simulation based on multiple sets of input wind data and the three-dimensional geometric structure; A plurality of measured wind data located in the three-dimensional geometric structure are extracted from the numerical simulation results, and their average value, i.e., the mean of the measured wind data, is calculated. A ratio of the mean of the measured wind data to the corresponding input wind data is calculated, and the preset wind change influence coefficient is determined based on the ratio corresponding to the plurality of the input wind data.
5. The automatic calibration method of a three-dimensional laser wind radar based on Transformer according to claim 1, Features: The work history data and the deviation tolerance coefficient are input into a deviation probability prediction model based on Transformer, and the deviation probability prediction model predicts the corresponding deviation probability, including: Sorting each group of wind measurement data in the work history data in the order of collection time, calculating the difference between adjacent wind measurement data after sorting, and obtaining a wind speed difference matrix and a wind direction difference matrix; Input each group of wind measurement data, the wind speed difference matrix, and the wind direction difference matrix into the deviation probability prediction model based on Transformer, the deviation probability prediction model predicts a first deviation probability, and uses the deviation tolerance coefficient to correct the first deviation probability to obtain a second deviation probability; The deviation probability prediction model outputs the second deviation probability.
6. The automatic calibration method of a three-dimensional laser wind radar based on Transformer according to claim 1, Features: The deviation tolerance factor is determined based on the usage information, including: A general large model is used to perform semantic understanding on the usage information, and a data accuracy evaluation is performed on the semantic understanding results based on network search content to determine the deviation tolerance coefficient.
7. An automatic calibration system for 3D laser wind radar based on Transformer. It is characterized in that Including acquisition module, deviation prediction module and automatic calibration module; The acquisition module is used to acquire the working history data of the three-dimensional laser wind measurement radar within a preset time period, wherein the working history data includes multiple groups of wind measurement data, wherein the wind measurement data includes wind speed data and wind direction data; and to acquire the usage information of the wind measurement data by the user, and determine the deviation tolerance coefficient according to the usage information; wherein the deviation tolerance coefficient is the degree of deviation between the wind measurement data and the actual wind data that the user can accept, and the greater the deviation degree, the greater the corresponding deviation tolerance coefficient, and vice versa; The deviation prediction module is used to input the work history data and the deviation tolerance coefficient into a deviation probability prediction model based on Transformer, and the deviation probability prediction model predicts the corresponding deviation probability; The automatic calibration module is used to control the laser wind measuring radar to perform automatic calibration in advance when the deviation probability is higher than the probability threshold.
