A method and system for dynamic phase calibration of a polarization-separated coherent wind-measuring lidar

Abstract

This invention discloses a phase dynamic calibration method and system for a polarization-separated coherent wind-measuring lidar, relating to the field of radar dynamic calibration technology. The method includes: acquiring individual phase calibration logs and operational logs, as well as co-source phase calibration logs and co-source operational logs of the target lidar, outputting sample phase calibration data; performing cross-commonality screening to obtain typical sample phase calibration data; performing multi-factor influence separation according to a preset factor separator, outputting the multi-factor influence separation results, and constructing multiple focused phase deviation analysis channels; integrating multiple focused phase deviation analysis channels with the separator to obtain a phase deviation dynamic accumulation component; acquiring the target lidar's lifecycle operational logs in real time and inputting them into the phase deviation dynamic accumulation component to obtain the accumulated phase deviation; and performing phase dynamic calibration based on the accumulated phase deviation. This invention solves the technical problem of poor radar dynamic calibration performance in existing technologies.

Description

Technical Field

[0001] This invention relates to the field of radar dynamic calibration technology, specifically to a phase dynamic calibration method and system for a polarization-separated coherent wind-measuring lidar. Background Technology

[0002] In the field of coherent wind lidar, polarization separation technology, by demodulating the orthogonal polarization components of atmospheric backscattered signals to achieve three-dimensional wind field inversion, has become a core method for high-precision wind field detection. However, this technology is extremely sensitive to system phase stability. Existing calibration schemes generally rely on static calibration strategies, and static parameters cannot dynamically adapt to cumulative changes, resulting in phase inaccuracies that severely restrict the long-term reliability and data validity of lidar. Summary of the Invention

[0003] This application provides a phase dynamic calibration method and system for a polarization-separated coherent wind-measuring lidar, which addresses the technical problem of inaccurate phase dynamic calibration of lidar in the prior art.

[0004] In view of the above problems, this application provides a phase dynamic calibration method and system for polarization-separated coherent wind lidar.

[0005] In a first aspect, this application provides a phase dynamic calibration method for a polarization-separated coherent wind-measuring lidar, the method comprising:

[0006] Acquire the individual phase calibration log, individual operation log, same-source phase calibration log, and same-source operation log of the target lidar, and output them as sample phase calibration data;

[0007] Cross-commonality screening is performed on the sample phase calibration data to obtain typical sample phase calibration data;

[0008] The typical sample phase calibration data are subjected to multi-factor influence separation according to the preset factor separator, and multiple single-influence factor sample data are output as multi-factor influence separation results. Multiple focused phase deviation analysis channels are constructed based on the multi-factor influence separation results.

[0009] Combined with the aforementioned separator, multiple focused phase deviation analysis channels are integrated to obtain a dynamic accumulation component for phase deviation.

[0010] The lifecycle operation log of the target radar is acquired in real time, and the lifecycle operation log is input into the phase deviation dynamic accumulation component for phase deviation accumulation evaluation to obtain the accumulated phase deviation.

[0011] Dynamic phase calibration is performed based on the accumulated phase deviation.

[0012] Secondly, this application provides a phase dynamic calibration system for a polarization-separated coherent wind-measuring lidar, comprising:

[0013] The data acquisition module is used to acquire individual phase calibration logs, individual operation logs, same-source phase calibration logs, and same-source operation logs of the target lidar, and outputs sample phase calibration data.

[0014] The data filtering module is used to perform cross-commonality filtering on the sample phase calibration data to obtain typical sample phase calibration data.

[0015] The influence separation module is used to perform multi-factor influence separation on the phase calibration data of the typical samples according to the preset factor separator, output multiple single influence factor sample data as multi-factor influence separation results, and construct multiple focused phase deviation analysis channels based on the multi-factor influence separation results.

[0016] The phase deviation analysis module is used in conjunction with the separator to integrate multiple focused phase deviation analysis channels to obtain the dynamic accumulation component of phase deviation.

[0017] The cumulative deviation analysis module is used to acquire the target radar's life cycle operation log in real time, and input the life cycle operation log into the phase deviation dynamic accumulation component to perform phase deviation accumulation evaluation and obtain the cumulative phase deviation.

[0018] A dynamic calibration module is used to perform dynamic phase calibration based on the accumulated phase deviation.

[0019] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0020] This application proposes a phase dynamic calibration method and system for polarization-separated coherent wind lidar. By constructing a multi-factor influence separation mechanism and a phase deviation dynamic accumulation module, the real-time performance and environmental adaptability of phase dynamic correction are significantly improved. Compared with traditional methods, the technical solution provided in this application significantly overcomes the limitations of static calibration.

[0021] This application achieves the technical effect of accurate and reliable dynamic phase compensation. Especially in complex scenarios, it significantly improves the long-term measurement reliability and data validity of radar systems, providing continuous and stable technical support for high-precision three-dimensional wind field detection. Attached Figure Description

[0022] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0023] Figure 1 This is a flowchart illustrating a phase dynamic calibration method for a polarization-separated coherent wind lidar provided in an embodiment of this application.

[0024] Figure 2 This is a schematic diagram of the phase dynamic calibration system of a polarization-separated coherent wind lidar provided in an embodiment of this application.

[0025] The components represented by each number in the attached diagram are explained below:

[0026] The system includes a data acquisition module 100, a data filtering module 200, an influence separation module 300, a phase deviation analysis module 400, a cumulative deviation analysis module 500, and a dynamic calibration module 600. Detailed Implementation

[0027] This application provides a phase dynamic calibration method and system for a polarization-separated coherent wind-measuring lidar, which addresses the technical problem of inaccurate phase dynamic calibration of lidar in the prior art.

[0028] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0029] It should be noted that the terms "comprising" and "having" are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or server that includes a series of steps or units is not necessarily limited to those steps or units that are explicitly listed, but may include other steps or modules that are not explicitly listed or that are inherent to these processes, methods, products, or devices.

[0030] Example 1, as Figure 1 As shown, this application provides a phase dynamic calibration method for a polarization-separated coherent wind lidar, wherein the method includes:

[0031] S10: Acquire the individual phase calibration log, individual operation log, same-source phase calibration log, and same-source operation log of the target lidar, and output them as sample phase calibration data.

[0032] The phase calibration of existing polarization-separated coherent wind lidar relies on discrete data sources. Individual logs only reflect the historical status of a single device, while logs from the same source lack horizontal correlation. This makes it difficult to extract universal patterns from the sample data, thus failing to support subsequent accurate modeling.

[0033] In step S10 of the method provided in this application embodiment, the work log includes at least the work temperature sequence, the work parameter sequence, and the work vibration sequence.

[0034] In this embodiment, by collecting historical logs of the target lidar, individual phase calibration logs and individual operation logs of the target lidar are obtained. Historical logs of other lidars originating from the same source as the target lidar are also collected to obtain co-origin phase calibration logs and co-origin operation logs. These individual phase calibration logs, individual operation logs, co-origin phase calibration logs, and co-origin operation logs are integrated and output as sample phase calibration data. The operation logs include at least the operation temperature sequence, operation parameter sequence, and operation vibration sequence.

[0035] By synchronously acquiring individual and source-related calibration logs and operation logs, correlated sample phase calibration data are constructed: First, individual logs ensure the basis for equipment-specific modeling; second, source-related logs provide a benchmark for group interference; and finally, the spatiotemporal alignment of operation logs and calibration logs makes the causal relationship between environmental parameters and phase response explicit, establishing a high-confidence data base for subsequent analysis.

[0036] S20: Perform cross-common screening on the sample phase calibration data to obtain typical sample phase calibration data.

[0037] The original sample data contains a large amount of atypical interference, such as sudden electromagnetic pulses and abnormal human operations, and the environmental parameters are unevenly distributed. If used directly for modeling, it will lead to two problems: firstly, outliers will distort the extraction of common patterns; secondly, the data in low-density areas lacks representativeness, resulting in weak model generalization ability. Traditional screening methods rely on only a single dimension and cannot solve the data confidence problem under multi-physics coupling.

[0038] Step S20 in the method provided in this application embodiment includes:

[0039] Based on the sample phase calibration data, and combined with the preset confidence constraints, a first constraint screening is performed to obtain the first screening result.

[0040] Based on the sample phase calibration data, a second constraint screening is performed in conjunction with a preset distribution density constraint to obtain the second screening result.

[0041] The intersection of the first screening result and the second screening result is determined, and the result is output as the phase calibration data of the typical sample.

[0042] In this embodiment, based on sample phase calibration data, a first constraint screening is performed using preset confidence constraints. The similarity between each sample phase calibration data and the mean of all sample phase calibration data is calculated and used as the confidence level. The confidence constraint is set according to the calibration accuracy requirements; the higher the accuracy requirement, the stricter the confidence constraint, and the higher the reliability of the first screening result. Similarity = 1 - |sample phase calibration data - mean of sample phase calibration data| ÷ [(sample phase calibration data + mean of sample phase calibration data) ÷ 2]. For example, if the preset confidence level is 80%, then sample phase calibration data with a similarity greater than or equal to 80% with the mean of sample phase calibration data are considered to meet the first constraint, and the first screening result is obtained.

[0043] Based on the sample phase calibration data, a second constraint screening is performed in conjunction with a preset distribution density constraint to obtain the second screening results. The distribution density constraint is set according to the calibration accuracy requirements. Higher accuracy requirements necessitate a stricter distribution density constraint, ensuring more representative second screening results. However, an overly strict constraint may result in too few qualified screening results; therefore, it cannot be set too low. For example, the distribution density of all sample phase calibration data is statistically analyzed, and a corresponding percentage density is selected as a threshold. This percentage can be determined by a professional technician based on the accuracy requirements. For instance, setting the distribution density constraint to 0.7, the data with a distribution density greater than 0.7 in the sample calibration data is calculated to obtain the second screening results.

[0044] Determine the intersection of the first and second screening results, and output the phase calibration data of typical samples.

[0045] This application employs a cross-similarity screening strategy, using confidence constraints to eliminate low-probability anomalies and distribution density constraints to ensure balanced coverage of all physical fields. The resulting typical samples possess both high purity and completeness, mitigating the risk of modeling distortion caused by data bias.

[0046] S30: Perform multi-factor influence separation on the phase calibration data of the typical sample according to the preset factor separator, output multiple single-influence factor sample data as multi-factor influence separation results, and construct multiple focused phase deviation analysis channels based on the multi-factor influence separation results.

[0047] Factors such as temperature, vibration, and time-varying characteristics are highly coupled in real-world environments. For example, thermal expansion can exacerbate mechanical deformation, and traditional univariate analysis methods may incorrectly attribute this to phase deviation. Forcibly fitting coupled data will mask the dominant role of the principal factors, ultimately leading to inaccurate prediction models.

[0048] Step S30 in the method provided in this application embodiment includes:

[0049] Based on the phase calibration data of the typical samples, the typical influencing factors of the target lidar are determined by principal component analysis.

[0050] The factor separator is initialized based on the typical influence factors, wherein the factor separator is constructed in conjunction with independent component analysis.

[0051] Input the typical sample phase calibration data into the initialized factor separator to perform multi-factor influence separation with the typical influence factor as the target, and obtain multiple sample data of the single influence factor.

[0052] The output of multiple sample data of the single influencing factor is the result of the multi-factor influence separation.

[0053] Traverse multiple single impact factor sample data to extract the corresponding single impact factor sequence datasets.

[0054] Using the cumulative rate of phase deviation as the dependent variable and the single influencing factor sequence data as the independent variable, multiple channels for focused phase deviation analysis are defined by combining an integral fitting strategy.

[0055] Supervised learning training is performed on each of the focused phase deviation analysis channels based on multiple single influence factor sequence datasets.

[0056] Among them, the multiple focused phase deviation analysis channels include at least:

[0057] The time-varying phase deviation analysis channel is used to predict the phase deviation of the target lidar caused by time variations based on the runtime.

[0058] The temperature variation-phase deviation analysis channel is used to predict the phase deviation of the target lidar caused by thermal effects based on the operating temperature.

[0059] The vibration-phase deviation analysis channel is used to predict the phase deviation of the target lidar caused by mechanical disturbances based on the structural vibration characteristics.

[0060] In this embodiment of the application, principal component analysis is performed using open-source and available principal component analysis tools, based on typical sample phase calibration data, to determine the typical influencing factors of the target lidar.

[0061] The factor separator is initialized based on typical impact factors. An independent component analysis model is constructed using the FastICA algorithm, which serves as the factor separator. Typical impact factors are set as the output dimension of the model, with a convergence threshold of 0.001 and a maximum number of iterations of 200, resulting in the initialized factor separator.

[0062] Input typical sample phase calibration data into the initialized factor separator to perform multi-factor influence separation targeting typical influence factors and obtain sample data of multiple single influence factors.

[0063] Output multiple single-factor sample data as the multi-factor influence separation result.

[0064] Traverse multiple single impact factor sample data to extract the corresponding single impact factor sequence datasets.

[0065] Using the cumulative rate of phase deviation as the dependent variable and the sequence data of a single influencing factor as the independent variable, multiple focused phase deviation analysis channels are defined using an integral fitting strategy. Each focused phase deviation analysis channel focuses on analyzing a single influencing factor, such as the effect of temperature variation. The cumulative error is estimated by integrating the error change rate, thus better reflecting the deviation evolution process during long-term system operation.

[0066] Based on multiple single-impact factor sequence datasets, supervised learning was performed on each focused phase bias analysis channel to obtain multiple focused phase bias analysis channels. The focused phase bias analysis channels are based on a "impact factor-bias cumulative rate" relationship model, where the cumulative bias is the result of integrating the model along the sequence data direction.

[0067] Among them, the multiple focused phase deviation analysis channels include at least:

[0068] The time-varying phase deviation analysis channel is used to predict the phase deviation of the target lidar caused by time variations based on the runtime.

[0069] The temperature variation-phase deviation analysis channel is used to predict the phase deviation of the target lidar caused by thermal effects based on the operating temperature.

[0070] The vibration-phase deviation analysis channel is used to predict the phase deviation of the target lidar caused by mechanical disturbance based on the structural vibration characteristics.

[0071] The factor separator identifies typical influencing factors through principal component analysis, decouples intertwined signals based on independent component analysis, reconstructs the independent action paths of each factor, extracts sample data of individual influencing factors, and eliminates cross-interference noise. Based on this, a dedicated phase deviation analysis channel is constructed, enabling accurate modeling of the deviation accumulation mechanisms of independent factors such as time-varying, temperature-varying, and vibration-related factors.

[0072] S40: Combine the separator with multiple focused phase deviation analysis channels to obtain the phase deviation dynamic accumulation component.

[0073] The separate operation of the analysis channels has two major limitations: first, it ignores the dynamic interaction between factors; second, the single channel output cannot establish a unified evaluation scale, resulting in a fragmented logic for calculating the cumulative phase deviation.

[0074] Step S40 in the method provided in this application embodiment includes:

[0075] Establish a mapping relationship between the typical influencing factors and multiple focused phase deviation analysis channels.

[0076] According to the mapping relationship, the multiple output terminals of the separator are connected to the input terminals of the multiple focused phase deviation analysis channels.

[0077] An integrated output layer is constructed and connected to the output terminals of multiple focused phase deviation analysis channels to obtain the phase deviation dynamic accumulation component.

[0078] In this embodiment, a mapping relationship is established between typical influence factors and multiple focused phase deviation analysis channels. The types of typical influence factors are the same as the types of focused phase deviation analysis.

[0079] Based on the mapping relationship, the corresponding output terminals of the connecting factor separator are connected to the input terminals of multiple focused phase deviation analysis channels.

[0080] By connecting the outputs of multiple channels focused on phase deviation analysis, an integrated output layer is constructed to obtain a dynamic phase deviation accumulation component.

[0081] By integrating multiple analysis channels using the mapping relationship of the factor separator, the output end of the separator is precisely connected to the input end of the channel to ensure that the data stream of each factor matches the dedicated analysis logic of the corresponding channel. The constructed phase deviation dynamic accumulation component, while retaining the advantages of independent factor analysis, realizes the collaborative evaluation of the coupling effect of multiple physical fields.

[0082] S50: Real-time acquisition of the target radar's lifecycle operation log, and input of the lifecycle operation log into the phase deviation dynamic accumulation component for phase deviation accumulation evaluation, thereby obtaining the accumulated phase deviation.

[0083] Static calibration models rely on extrapolation from historical data and cannot respond to real-time changes in operating conditions.

[0084] In this embodiment, the operational log of the target radar during its current lifecycle is acquired in real time. The operational log includes at least an operational temperature sequence, an operational parameter sequence, and an operational vibration sequence. The lifecycle operational log is input to a phase deviation dynamic accumulation component for phase deviation accumulation evaluation, thereby obtaining the accumulated phase deviation.

[0085] In the solution provided in this application, the lifecycle operation log is input to the dynamic accumulation module in real time, so that the phase deviation evaluation is strictly synchronized with the real environment.

[0086] S60: Perform dynamic phase calibration based on the accumulated phase deviation.

[0087] Static compensation strategies will further amplify the system phase tolerance.

[0088] In this embodiment, phase dynamic calibration is performed based on the cumulative phase deviation. Specifically, a constant step size method is used for phase dynamic calibration. For example, the step size is set to the cumulative phase deviation / 2 to perform phase dynamic calibration, ensuring that the system phase stability continues to converge under all operating conditions.

[0089] Example 2, as Figure 2 As shown, based on the same inventive concept as the phase dynamic calibration method for a polarization-separated coherent wind lidar provided in Embodiment 1, this embodiment of the invention also provides a phase dynamic calibration system for a polarization-separated coherent wind lidar, comprising:

[0090] The data acquisition module 100 is used to acquire the individual phase calibration log, individual operation log, same-source phase calibration log, and same-source operation log of the target lidar, and outputs sample phase calibration data.

[0091] The data filtering module 200 is used to perform cross-commonality filtering on the sample phase calibration data to obtain typical sample phase calibration data.

[0092] The influence separation module 300 is used to perform multi-factor influence separation on the phase calibration data of the typical sample according to the preset factor separator, output multiple single influence factor sample data as multi-factor influence separation results, and construct multiple focused phase deviation analysis channels based on the multi-factor influence separation results.

[0093] The phase deviation analysis module 400 is used in conjunction with the separator to integrate multiple focused phase deviation analysis channels to obtain the phase deviation dynamic accumulation component.

[0094] The cumulative deviation analysis module 500 is used to acquire the life cycle operation log of the target radar in real time, and input the life cycle operation log into the phase deviation dynamic accumulation component to perform phase deviation accumulation evaluation and obtain the cumulative phase deviation.

[0095] The dynamic calibration module 600 is used to perform dynamic phase calibration based on the accumulated phase deviation.

[0096] In one embodiment, the data acquisition module 100 further includes a work log that includes at least a work temperature sequence, a work parameter sequence, and a work vibration sequence.

[0097] In one embodiment, the data filtering module 200 is further configured to:

[0098] Based on the sample phase calibration data, and combined with the preset confidence constraints, a first constraint screening is performed to obtain the first screening result.

[0099] Based on the sample phase calibration data, a second constraint screening is performed in conjunction with a preset distribution density constraint to obtain the second screening result.

[0100] The intersection of the first screening result and the second screening result is determined, and the result is output as the phase calibration data of the typical sample.

[0101] In one embodiment, the separation module 300 is further configured to:

[0102] Based on the phase calibration data of the typical samples, the typical influencing factors of the target lidar are determined by principal component analysis.

[0103] The factor separator is initialized based on the typical influence factors, wherein the factor separator is constructed in conjunction with independent component analysis.

[0104] Input the typical sample phase calibration data into the initialized factor separator to perform multi-factor influence separation with the typical influence factor as the target, and obtain multiple sample data of the single influence factor.

[0105] The output of multiple sample data of the single influencing factor is the result of the multi-factor influence separation.

[0106] Traverse multiple single impact factor sample data to extract the corresponding single impact factor sequence datasets.

[0107] Using the cumulative rate of phase deviation as the dependent variable and the single influencing factor sequence data as the independent variable, multiple focused phase deviation analysis channels are defined by combining an integral fitting strategy.

[0108] Supervised learning training is performed on each of the focused phase deviation analysis channels based on multiple single influence factor sequence datasets.

[0109] Among them, the multiple focused phase deviation analysis channels include at least:

[0110] The time-varying phase deviation analysis channel is used to predict the phase deviation of the target lidar caused by time variations based on the runtime.

[0111] The temperature variation-phase deviation analysis channel is used to predict the phase deviation of the target lidar caused by thermal effects based on the operating temperature.

[0112] The vibration-phase deviation analysis channel is used to predict the phase deviation of the target lidar caused by mechanical disturbances based on the structural vibration characteristics.

[0113] In one embodiment, the phase deviation analysis module 400 is further configured to:

[0114] Establish a mapping relationship between the typical influencing factors and multiple focused phase deviation analysis channels.

[0115] According to the mapping relationship, the multiple output terminals of the separator are connected to the input terminals of the multiple focused phase deviation analysis channels.

[0116] An integrated output layer is constructed and connected to the output terminals of multiple focused phase deviation analysis channels to obtain the phase deviation dynamic accumulation component.

[0117] In summary, the embodiments of this application have at least the following technical effects:

[0118] This application proposes a phase dynamic calibration method and system for polarization-separated coherent wind lidar. By constructing a multi-factor influence separation mechanism and a phase deviation dynamic accumulation module, the real-time performance and environmental adaptability of phase dynamic correction are significantly improved. Compared with traditional methods, the technical solution provided in this application significantly overcomes the limitations of static calibration: First, the typical sample extraction strategy based on cross-commonality screening effectively extracts data features and avoids redundant noise interference; second, the factor separator's decoupling capability for the influence of multiple physical fields such as temperature, vibration, and time-varying factors enables accurate modeling of the phase deviation evolution law of each independent factor; and third, by integrating a dynamic accumulation module with multiple focus analysis channels, synchronous tracking of transient disturbances and long-term drift in the operating environment is achieved, allowing the phase compensation parameters to evolve autonomously with the actual radar operating conditions. Especially in complex scenarios, this significantly improves the long-term measurement reliability and data validity of the radar system, providing continuous and stable technical support for high-precision three-dimensional wind field detection.

[0119] This application achieves the technical effect of accurate and reliable phase dynamic compensation.

[0120] It should be noted that the order of the embodiments described above is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. Furthermore, the above description focuses on specific embodiments of this specification. Additionally, the processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some implementations, multitasking and parallel processing are possible or may be advantageous.

[0121] The above description is only a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

[0122] This specification and accompanying drawings are merely illustrative examples of this application and are intended to cover any and all modifications, variations, combinations, or equivalents within the scope of this application. Clearly, those skilled in the art can make various alterations and modifications to this application without departing from its scope. Therefore, if such modifications and modifications fall within the scope of this application and its equivalents, this application intends to include such modifications and modifications.

Claims

1. A phase dynamic calibration method for a polarization-separated coherent wind-measuring lidar, characterized in that, include: Acquire the individual phase calibration log, individual operation log, same-source phase calibration log, and same-source operation log of the target lidar, and output them as sample phase calibration data; Cross-commonality screening is performed on the sample phase calibration data to obtain typical sample phase calibration data; The typical sample phase calibration data are subjected to multi-factor influence separation according to the preset factor separator, and multiple single-influence factor sample data are output as multi-factor influence separation results. Multiple focused phase deviation analysis channels are constructed based on the multi-factor influence separation results. By combining the aforementioned factor separator with multiple focused phase deviation analysis channels, a dynamic accumulation component for phase deviation can be obtained. The lifecycle operation log of the target radar is acquired in real time, and the lifecycle operation log is input into the phase deviation dynamic accumulation component for phase deviation accumulation evaluation to obtain the accumulated phase deviation. Dynamic phase calibration is performed based on the accumulated phase deviation.

2. The phase dynamic calibration method for a polarization-separated coherent wind lidar as described in claim 1, characterized in that, The operation log should include at least the operation temperature sequence, operation parameter sequence, and operation vibration sequence.

3. The phase dynamic calibration method for a polarization-separated coherent wind lidar as described in claim 2, characterized in that, Cross-commonality screening is performed on the sample phase calibration data to obtain typical sample phase calibration data, including: Based on the sample phase calibration data, and combined with the preset confidence constraints, a first constraint screening is performed to obtain the first screening result; Based on the sample phase calibration data, a second constraint screening is performed in conjunction with a preset distribution density constraint to obtain the second screening result; The intersection of the first screening result and the second screening result is determined, and the result is output as the phase calibration data of the typical sample.

4. The phase dynamic calibration method for a polarization-separated coherent wind lidar as described in claim 3, characterized in that, The typical sample phase calibration data are subjected to multi-factor influence separation according to a preset factor separator, and the multi-factor influence separation results are output as multiple single-influence factor sample data, including: Based on the phase calibration data of the typical samples, the typical influencing factors of the target lidar are determined by principal component analysis; The factor separator is initialized based on the typical influence factors, wherein the factor separator is constructed in conjunction with independent component analysis; Input the typical sample phase calibration data into the initialized factor separator to perform multi-factor influence separation with the typical influence factor as the target, and obtain multiple single influence factor sample data; The output of multiple sample data of the single influencing factor is the result of the multi-factor influence separation.

5. The phase dynamic calibration method for a polarization-separated coherent wind lidar as described in claim 4, characterized in that, Based on the results of the separation of multi-factor influences, multiple focused phase bias analysis channels were constructed, including: Traverse multiple single impact factor sample data and extract the corresponding multiple single impact factor sequence datasets; Using the cumulative rate of phase deviation as the dependent variable and the single influencing factor sequence data as the independent variable, multiple focused phase deviation analysis channels are defined by combining an integral fitting strategy. Supervised learning training is performed on each of the focused phase deviation analysis channels based on multiple single influence factor sequence datasets.

6. The phase dynamic calibration method for a polarization-separated coherent wind lidar as described in claim 1, characterized in that, The multiple focused phase deviation analysis channels mentioned above include at least: The time-varying phase deviation analysis channel is used to predict the phase deviation of the target lidar caused by time variation based on the runtime. The temperature variation-phase deviation analysis channel is used to predict the phase deviation of the target lidar caused by thermal effects based on the operating temperature. The vibration-phase deviation analysis channel is used to predict the phase deviation of the target lidar caused by mechanical disturbance based on the structural vibration characteristics.

7. The phase dynamic calibration method for a polarization-separated coherent wind lidar as described in claim 5, characterized in that, In conjunction with the aforementioned factor separator, multiple focused phase deviation analysis channels are integrated to obtain a dynamic accumulation component for phase deviation, including: Establish a mapping relationship between the typical influencing factors and multiple focused phase deviation analysis channels; According to the mapping relationship, the multiple output terminals of the factor separator are connected to the input terminals of the multiple focused phase deviation analysis channels; An integrated output layer is constructed and connected to the output terminals of multiple focused phase deviation analysis channels to obtain the phase deviation dynamic accumulation component.

8. A phase dynamic calibration system for a polarization-separated coherent wind-measuring lidar, characterized in that, A phase dynamic calibration method for implementing the polarization-separated coherent wind lidar according to any one of claims 1 to 7, the system comprising: The data acquisition module is used to acquire the individual phase calibration log, individual operation log, same-source phase calibration log, and same-source operation log of the target lidar, and outputs sample phase calibration data. The data filtering module is used to perform cross-commonality filtering on the sample phase calibration data to obtain typical sample phase calibration data; The influence separation module is used to perform multi-factor influence separation on the phase calibration data of the typical samples according to the preset factor separator, output multiple single influence factor sample data as multi-factor influence separation results, and construct multiple focused phase deviation analysis channels based on the multi-factor influence separation results; The phase deviation analysis module is used to combine the factor separator with multiple focused phase deviation analysis channels to obtain the phase deviation dynamic accumulation component. The cumulative deviation analysis module is used to acquire the target radar's life cycle operation log in real time, and input the life cycle operation log into the phase deviation dynamic accumulation component to perform phase deviation accumulation evaluation and obtain the cumulative phase deviation. A dynamic calibration module is used to perform dynamic phase calibration based on the accumulated phase deviation.

Images