Temperature and pressure synchronous sensing method and device for monitoring parameters of bridge ship collision prevention system

Abstract

The invention provides a temperature and pressure synchronous sensing method and device for monitoring parameters of a bridge ship collision prevention system, and relates to the technical field of optical fiber sensing. The method comprises a sensor calibration module and a real-time demodulation module. The sensor calibration module comprises a sensor calibration unit, a spectral feature extraction unit, a database unit, a data preprocessing and standardization unit, a model construction and training unit and a hyper-parameter optimization unit. The model building and training unit is used for building a multi-output support vector regression model, specifically, regression functions are built for temperature and pressure respectively, and meanwhile a first SVR model and a second SVR model are included. According to the invention, synchronous and high-precision measurement of temperature and pressure can be realized by using a single sensing unit and a single demodulation model; therefore, the reliability, the compactness and the long-term stability of the system are remarkably improved while the measurement precision is ensured, and the harsh requirements for safety monitoring of major infrastructures such as a ship colliding with a bridge and the like are met.

Description

Technical Field

[0001] This invention relates to the field of fiber optic sensing technology, and in particular to a method and device for synchronously sensing temperature and pressure for monitoring parameters of a bridge anti-ship collision system. Background Technology

[0002] Bridge collision avoidance monitoring systems need to sense the impact force (pressure) of ships and changes in ambient temperature in real time, as both cause stress and strain in the bridge structure. However, their physical mechanisms and effects on the structure are fundamentally different and must be distinguished. A hybrid interference structure composed of single-mode fiber, hollow fiber, and coreless fiber fusion splicing can generate an interference spectrum that is highly sensitive to external physical quantities, possessing the potential for ultra-compact, high-sensitivity sensing. It is well-suited for embedding in bridge protection facilities or critical stress-bearing components to monitor local pressure distribution and long-term temperature field changes at the moment of a ship collision. However, this structure faces a core technical bottleneck in practical applications: its interference spectrum is simultaneously affected by the cross-sensitivity of temperature and pressure. A change in one physical quantity can be "disguised" as a signal of another, causing the system to be unable to determine whether the pressure change is due to a ship collision or merely a normal fluctuation in ambient temperature. This fundamental risk of misjudgment severely hinders the practical application of this type of sensor in bridge safety monitoring, which requires high reliability. To address the cross-sensitivity problem, the most typical solution in existing technologies is to adopt a strategy of "dual-channel sensing unit + discrete algorithm demodulation." Specifically, a common approach is to deploy two sensors: a main sensor sensitive to both temperature and pressure, and a temperature-compensated sensor placed in a reference environment, sensitive only to temperature. The system needs to build demodulation models for both sensors separately. First, the temperature value is calculated using the reference sensor, then substituted into the main sensor's model to subtract the temperature effect, finally calculating the pressure value. A similar technique involves fabricating two interferometers with different sensitivities to temperature and pressure within a single sensor, but its backend still relies on two independently constructed demodulation models for step-by-step calculations. While these existing solutions alleviate the problem to some extent, they also introduce significant drawbacks: First, they rely on multiple sensors or complex sensing structures, increasing system cost, packaging complexity, and failure rate, reducing the feasibility of large-scale deployment on bridges. Second, the step-by-step demodulation calculation process is lengthy, not only increasing the burden on the signal processing unit but also accumulating and propagating errors in the two calculation steps, ultimately limiting the absolute accuracy and reliability of pressure measurement. For applications like ship-bridge collision monitoring, which require instantaneous capture of weak collision signals and elimination of interference from diurnal temperature variations, existing technologies struggle to balance demodulation accuracy, system complexity, and real-time performance. A more direct, efficient, and accurate technical solution is urgently needed. While temperature-compensated pressure measurement systems based on reference sensors theoretically offer a solution to cross-sensitivity, in practical engineering applications, especially in ship-bridge collision monitoring scenarios where reliability and accuracy are paramount, they exhibit several inherent and insurmountable drawbacks. Firstly, the system relies on dual sensing units, directly leading to a significant increase in hardware costs, a substantial increase in system packaging complexity, and an overall higher failure rate.In the harsh service environment of bridges, the failure of any single sensor will lead to the loss of the entire monitoring function, severely testing the system's reliability. Secondly, its step-by-step demodulation algorithm architecture introduces significant error accumulation and propagation effects into the signal processing chain. Any tiny error in the first step of temperature calculation will be amplified in the second step of pressure calculation, severely affecting the absolute accuracy of the final pressure measurement result and potentially leading to misjudgment of the ship's impact force. Furthermore, to achieve pressure isolation, the special packaging process of the reference sensor is complex and it is difficult to guarantee its absolute stability under long-term vibration and humid heat aging, introducing uncontrollable uncertainties into the measurement. Finally, this scheme requires the establishment and maintenance of two independent demodulation models, which increases the workload and complexity of system calibration and raises the computational resource requirements of the signal processing unit. These shortcomings collectively restrict the widespread application of this technology in bridge safety monitoring, which requires high reliability, high accuracy, and large-scale deployment. Summary of the Invention

[0003] To address the technical problem that reliance on dual sensing units in existing technologies leads to system complexity and vulnerability, this invention provides a method and apparatus for synchronously sensing temperature and pressure to monitor parameters of a bridge anti-ship collision system. The technical solution is as follows:

[0004] On the one hand, a method for synchronously sensing temperature and pressure to monitor parameters of a bridge anti-ship collision system is provided. This method is implemented by a synchronous sensing device for temperature and pressure monitoring of bridge anti-ship collision system parameters, including a sensor calibration module and a real-time demodulation module. Its characteristic is that it includes:

[0005] The sensor calibration module includes a sensor calibration unit, a spectral feature extraction unit, a database unit, a data preprocessing and standardization unit, a model building and training unit, and a hyperparameter optimization unit. The sensor calibration unit is used to collect raw spectral data. The model building and training unit is used to build a multi-output support vector regression model. The construction of the multi-output support vector regression model includes establishing regression functions for temperature and pressure respectively, so that the model learns the independent mapping relationship between temperature and pressure from the same set of spectral features. The multi-output support vector regression model includes a first SVR model and a second SVR model.

[0006] The real-time demodulation module includes a data acquisition unit, a preprocessing unit, a standardization unit, and a model inference unit.

[0007] Preferably, the sensor calibration unit includes:

[0008] The prepared single-mode hollow-coreless fiber optic sensor was placed in a high-precision temperature and pressure control chamber.

[0009] A sensor calibration task instruction is generated to perform the calibration process and obtain raw spectral data. The sensor calibration task instruction includes defining measurement conditions based on the expected application range, setting calibration points, and for each combination of temperature and pressure values, collecting complete spectral data multiple times using a spectrometer with a spectral resolution of not less than 0.01 nm. The multiple collection of complete spectral data includes collecting multiple samples at each calibration point to reduce the influence of random noise. The calibration points correspond to a series of temperature and pressure value combinations generated within the expected application range.

[0010] Preferably, the spectral feature extraction unit includes:

[0011] Based on the original spectral data, the main interference valley wavelength is extracted from each original spectrum. The extraction of the main interference valley wavelength from each original spectrum includes: accurately locating it using cubic spline interpolation. The main interference valley wavelength corresponds to the wavelength value corresponding to the lowest transmittance in the spectrum.

[0012] Based on the original spectral data, the wavelengths of secondary interference valleys / peaks are extracted from each original spectrum. The extraction of the wavelengths of secondary interference valleys / peaks from each original spectrum includes: obtaining them by finding local extreme points. The wavelength of the secondary interference valley / peak is the wavelength value of another significant interference extreme point, which is a valley or a peak.

[0013] Based on the original spectral data, the depth of the main interference valley is calculated from each original spectrum. The calculation of the depth of the main interference valley from each original spectrum includes: calculating based on the light intensity at the bottom of the valley. The depth of the main interference valley is the normalized light intensity value at the bottom of the main interference valley.

[0014] Based on the original spectral data, the second-order central moment of the spectrum is calculated from each original spectrum. The calculation of the second-order central moment of the spectrum from each original spectrum includes: calculation based on the wavelength of the spectral centroid and the light intensity of the corresponding wavelength. The second-order central moment of the spectrum characterizes the broadening of the spectral energy distribution.

[0015] Based on the original spectral data, the spectral asymmetry is calculated from each original spectrum. The calculation of the spectral asymmetry from each original spectrum includes: calculation based on the spectral centroid wavelength, the light intensity at the corresponding wavelength, and the second central moment. The spectral asymmetry is used to quantify the asymmetry of the spectral distribution.

[0016] A multidimensional feature vector is formed based on the wavelength of the primary interference valley, the wavelength of the secondary interference valley / peak, the depth of the primary interference valley, the second-order central moment of the spectrum, and the spectral asymmetry.

[0017] Preferably, the database unit includes:

[0018] At each calibration point, the corresponding spectral feature vector is extracted based on the spectral feature extraction unit;

[0019] Synchronously record the corresponding real physical quantity labels;

[0020] Each calibration point, its corresponding feature vector, and its corresponding physical quantity label are assembled into a sample group and added to the database. The physical quantity labels include temperature labels and pressure labels.

[0021] The database is then subjected to quality checks and annotation improvements to obtain a complete database.

[0022] Preferably, the model building and training unit includes:

[0023] A nonlinear mapping is used to map the input spectral feature vector to a high-dimensional feature space;

[0024] Two optimal linear regression hyperplanes for temperature and pressure are constructed in this space respectively;

[0025] Using the same input feature vector, two mapping functions are constructed to train independent SVR models for temperature and pressure, respectively corresponding to the first SVR model and the second SVR model, resulting in a multi-output support vector regression model. The first SVR model includes a mapping function for temperature prediction, which includes the weight vector and bias term of the temperature model. The second regression model includes a mapping function for pressure prediction, which includes the weight vector and bias term of the pressure model.

[0026] Preferably, the hyperparameter optimization unit includes:

[0027] Based on the complete database, extract spectral feature vectors;

[0028] The spectral feature vectors are transformed into standardized input features using a standardization process.

[0029] Using standardized input features as input, and temperature and pressure labels as targets, the first and second SVR models are trained respectively.

[0030] A strategy combining cross-validation and grid search is used to optimize the hyperparameters of the first SVR model and the second SVR model. The cross-validation strategy involves randomly dividing the calibration dataset into mutually exclusive subsets, then using the first subset of each mutually exclusive subset as the validation set and the first subset as the training set, training the model on the training set, and evaluating its prediction performance on the validation set. The grid search strategy involves traversing the hyperparameter combination space defined by the grid.

[0031] After model training is completed, the performance of the first SVR model and the second SVR model is comprehensively evaluated using an independent test set and preset evaluation metrics.

[0032] Preferably, the real-time demodulation module includes a data acquisition unit, a preprocessing unit, a standardization unit, and a model inference unit, comprising:

[0033] After the sensor starts the measurement, the data acquisition unit immediately acquires the raw spectrum in real time;

[0034] The preprocessing unit extracts feature vectors from the raw spectrum;

[0035] The standardization unit performs standardization processing on the extracted feature vectors to obtain standardized feature vectors.

[0036] The model inference unit inputs the standardized feature vectors into the pre-trained multi-output support vector regression model to obtain the predicted values ​​of temperature and pressure.

[0037] The model output results are checked for reasonableness, including outlier detection. If an outlier is detected, a remeasurement mechanism is triggered.

[0038] On the other hand, a temperature and pressure synchronous sensing device for monitoring parameters of a bridge anti-ship collision system is provided. This device is applied to the temperature and pressure synchronous sensing method for monitoring parameters of a bridge anti-ship collision system. The device includes:

[0039] Sensor calibration module: The sensor calibration module includes a sensor calibration unit, a spectral feature extraction unit, a database unit, a data preprocessing and standardization unit, a model building and training unit, and a hyperparameter optimization unit. The sensor calibration unit is used to collect raw spectral data. The model building and training unit is used to build a multi-output support vector regression model. The construction of the multi-output support vector regression model includes establishing regression functions for temperature and pressure respectively, so that the model learns the independent mapping relationship between temperature and pressure from the same set of spectral features. The multi-output support vector regression model includes a first SVR model and a second SVR model.

[0040] Real-time demodulation module: The real-time demodulation module includes a data acquisition unit, a preprocessing unit, a standardization unit, and a model inference unit.

[0041] On the other hand, a temperature and pressure synchronous sensing device for monitoring parameters of a bridge anti-ship collision system is provided. The temperature and pressure synchronous sensing device for monitoring parameters of a bridge anti-ship collision system includes: a processor; a memory, wherein the memory stores computer-readable instructions, and when the computer-readable instructions are executed by the processor, the method described in any one of the above-described methods for synchronous sensing of temperature and pressure for monitoring parameters of a bridge anti-ship collision system is implemented.

[0042] On the other hand, a computer-readable storage medium is provided, characterized in that the computer-readable storage medium stores program code, which can be invoked by a processor to execute the method as described in any one of claims 1 to 7.

[0043] The beneficial effects of the technical solutions provided in the embodiments of the present invention include at least the following:

[0044] This scheme delves into the synergistic variation patterns of multiple features in the output spectrum of fiber optic sensors based on a hybrid cascaded structure of single-mode, hollow-core, and coreless fibers. By constructing a single support vector regression model, it achieves synchronous and high-precision demodulation of temperature and pressure signals. This method fundamentally abandons the traditional dual-sensor or dual-channel demodulation architecture, effectively distinguishing temperature and pressure signals using only a single, low-cost model. It significantly simplifies the system structure, reduces packaging difficulty and cost, and provides a novel technical approach to improve the reliability of ship-bridge collision monitoring. Attached Figure Description

[0045] 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.

[0046] Figure 1 This is a schematic diagram of a synchronous temperature and pressure sensing system for monitoring parameters of a bridge anti-ship collision system provided in an embodiment of the present invention;

[0047] Figure 2 This is a flowchart of a method for synchronously sensing temperature and pressure to monitor parameters of a bridge anti-ship collision system provided in an embodiment of the present invention;

[0048] Figure 3 This is a block diagram of a temperature and pressure synchronous sensing device for monitoring parameters of a bridge anti-ship collision system provided in an embodiment of the present invention;

[0049] Figure 4 This is a schematic diagram of the structure of a synchronous temperature and pressure sensing device for monitoring parameters of a bridge anti-ship collision system provided in an embodiment of the present invention. Detailed Implementation

[0050] The technical solution of the present invention will now be described with reference to the accompanying drawings.

[0051] In embodiments of the present invention, words such as "exemplarily," "for example," etc., are used to indicate that something is an example, illustration, or description. Any embodiment or design described as "exemplary" in the present invention should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of the word "exemplary" is intended to present the concept in a concrete manner. Furthermore, in embodiments of the present invention, the meaning expressed by "and / or" can be both, or either one.

[0052] In the embodiments of this invention, the terms "image" and "picture" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, they convey the same meaning. Similarly, the terms "of," "corresponding (relevant)," and "corresponding" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, they convey the same meaning.

[0053] In this embodiment of the invention, sometimes a subscript such as W1 may be written in a non-subscript form such as W1. When the difference is not emphasized, the meaning they express is the same.

[0054] To make the technical problems, technical solutions and advantages of the present invention clearer, a detailed description will be given below in conjunction with the accompanying drawings and specific embodiments.

[0055] This invention provides a method for synchronously sensing temperature and pressure to monitor parameters of a bridge anti-ship collision system. This method can be implemented using a synchronous sensing device for temperature and pressure monitoring of the bridge anti-ship collision system parameters, which can be a terminal or a server. Figure 1 The flowchart shown is a method for synchronously sensing temperature and pressure to monitor parameters of a bridge anti-ship collision system. The processing flow of this method may include the following steps:

[0056] The sensor calibration module includes a sensor calibration unit, a spectral feature extraction unit, a database unit, a data preprocessing and standardization unit, a model building and training unit, and a hyperparameter optimization unit. The sensor calibration unit is used to collect raw spectral data. The model building and training unit is used to build a multi-output support vector regression model. The construction of the multi-output support vector regression model includes establishing regression functions for temperature and pressure respectively, so that the model learns the independent mapping relationship between temperature and pressure from the same set of spectral features. The multi-output support vector regression model includes a first SVR model and a second SVR model.

[0057] Preferably, the sensor calibration unit includes:

[0058] The prepared single-mode hollow-coreless fiber optic sensor was placed in a high-precision temperature and pressure control chamber.

[0059] A sensor calibration task instruction is generated to perform the calibration process and obtain raw spectral data. The sensor calibration task instruction includes defining measurement conditions based on the expected application range, setting calibration points, and for each combination of temperature and pressure values, collecting complete spectral data multiple times using a spectrometer with a spectral resolution of not less than 0.01 nm. The multiple collection of complete spectral data includes collecting multiple samples at each calibration point to reduce the influence of random noise. The calibration points correspond to a series of temperature and pressure value combinations generated within the expected application range.

[0060] Preferably, the spectral feature extraction unit includes:

[0061] Based on the original spectral data, the main interference valley wavelength is extracted from each original spectrum. The extraction of the main interference valley wavelength from each original spectrum includes: accurately locating it using cubic spline interpolation. The main interference valley wavelength corresponds to the wavelength value corresponding to the lowest transmittance in the spectrum.

[0062] Based on the original spectral data, the wavelengths of secondary interference valleys / peaks are extracted from each original spectrum. The extraction of the wavelengths of secondary interference valleys / peaks from each original spectrum includes: obtaining them by finding local extreme points. The wavelength of the secondary interference valley / peak is the wavelength value of another significant interference extreme point, which is a valley or a peak.

[0063] Based on the original spectral data, the depth of the main interference valley is calculated from each original spectrum. The calculation of the depth of the main interference valley from each original spectrum includes: calculating based on the light intensity at the bottom of the valley. The depth of the main interference valley is the normalized light intensity value at the bottom of the main interference valley.

[0064] Based on the original spectral data, the second-order central moment of the spectrum is calculated from each original spectrum. The calculation of the second-order central moment of the spectrum from each original spectrum includes: calculation based on the wavelength of the spectral centroid and the light intensity of the corresponding wavelength. The second-order central moment of the spectrum characterizes the broadening of the spectral energy distribution.

[0065] Based on the original spectral data, the spectral asymmetry is calculated from each original spectrum. The calculation of the spectral asymmetry from each original spectrum includes: calculation based on the spectral centroid wavelength, the light intensity at the corresponding wavelength, and the second central moment. The spectral asymmetry is used to quantify the asymmetry of the spectral distribution.

[0066] A multidimensional feature vector is formed based on the wavelength of the primary interference valley, the wavelength of the secondary interference valley / peak, the depth of the primary interference valley, the second-order central moment of the spectrum, and the spectral asymmetry.

[0067] Preferably, the database unit includes:

[0068] At each calibration point, the corresponding spectral feature vector is extracted based on the spectral feature extraction unit;

[0069] Synchronously record the corresponding real physical quantity labels;

[0070] Each calibration point, its corresponding feature vector, and its corresponding physical quantity label are assembled into a sample group and added to the database. The physical quantity labels include temperature labels and pressure labels.

[0071] The database is then subjected to quality checks and annotation improvements to obtain a complete database.

[0072] Preferably, the model building and training unit includes:

[0073] A nonlinear mapping is used to map the input spectral feature vector to a high-dimensional feature space;

[0074] Two optimal linear regression hyperplanes for temperature and pressure are constructed in this space respectively;

[0075] Using the same input feature vector, two mapping functions are constructed to train independent SVR models for temperature and pressure, respectively corresponding to the first SVR model and the second SVR model, resulting in a multi-output support vector regression model. The first SVR model includes a mapping function for temperature prediction, which includes the weight vector and bias term of the temperature model. The second regression model includes a mapping function for pressure prediction, which includes the weight vector and bias term of the pressure model.

[0076] Preferably, the hyperparameter optimization unit includes:

[0077] Based on the complete database, extract spectral feature vectors;

[0078] The spectral feature vectors are transformed into standardized input features using a standardization process.

[0079] Using standardized input features as input, and temperature and pressure labels as targets, the first and second SVR models are trained respectively.

[0080] A strategy combining cross-validation and grid search is used to optimize the hyperparameters of the first SVR model and the second SVR model. The cross-validation strategy involves randomly dividing the calibration dataset into mutually exclusive subsets, then using the first subset of each mutually exclusive subset as the validation set and the first subset as the training set, training the model on the training set, and evaluating its prediction performance on the validation set. The grid search strategy involves traversing the hyperparameter combination space defined by the grid.

[0081] After model training is completed, the performance of the first SVR model and the second SVR model is comprehensively evaluated using an independent test set and preset evaluation metrics.

[0082] In some embodiments, the prepared single-mode hollow-coreless fiber optic sensor is placed in a high-precision temperature and pressure control chamber, within a temperature range covering the intended application area. and pressure value In combination, the corresponding output interference spectra are measured and recorded one by one. This calibration process requires ensuring the accuracy of temperature and pressure control; typically, the temperature control accuracy should reach ±0.1℃, and the pressure control accuracy should reach ±0.1%FS. For each For the combined data, a spectrometer with a spectral resolution of at least 0.01 nm is required to collect complete spectral data. Multiple samples should be collected at each calibration point to reduce the impact of random noise.

[0083] Next, a set of key features that comprehensively reflect the morphological changes are extracted from each raw spectrum. These features are carefully selected and include multiple dimensions with different sensitivities to temperature and pressure:

[0084] The wavelength of the main interference valley ( The wavelength value corresponding to the lowest transmittance in the spectrum is accurately located using cubic spline interpolation. It is sensitive to changes in temperature and pressure, but the sensitivity coefficients differ.

[0085] Secondary interference valley / peak wavelength ( Another significant interference extremum point has a wavelength value obtained by finding local extrema. Its sensitivity coefficient to temperature and pressure is related to... Significant differences exist, and these differences provide an important basis for the model to distinguish the effects of temperature and pressure.

[0086] Main interference valley depth ( The normalized light intensity value at the bottom of the main interference valley is calculated as follows: ,in The light intensity is at the bottom of the valley. As a reference light intensity, it is mainly affected by loss modulation and has unique response characteristics to changes in physical quantities.

[0087] Second central moment of the spectrum ( ): Characterizes the broadening of the spectral energy distribution, calculated as ,in For wavelength, The wavelength of the spectral centroid Light intensity.

[0088] This feature is sensitive to the overall deformation of the spectrum, providing overall morphological information that differs from the location of extreme points.

[0089] Spectral asymmetry ( ): Calculated as ,in The standard deviation is used to quantify the asymmetry of the spectral distribution, a characteristic that has a unique response to certain specific temperature-pressure combinations.

[0090] This step constructs a database containing M×N samples, each sample consisting of a set of feature vectors. and corresponding real physical quantity labels This database not only contains rich spectral feature information, but more importantly, it records the variation patterns of these features under different combinations of temperature and pressure, laying a solid foundation for subsequent training of intelligent models capable of understanding the complex nonlinear relationships between spectral features and temperature and pressure.

[0091] Support vector regression is a powerful regression method based on statistical learning theory. Its core idea is to map data to a high-dimensional feature space through nonlinear mapping and construct the optimal linear regression hyperplane in this space.

[0092] To achieve temperature ( ) and pressure ( For simultaneous prediction of output variables, this invention employs a multi-output support vector regression method. The core idea of ​​this method is to establish an independent regression model for each output variable, while simultaneously considering the correlation between outputs. Specifically, for each output variable (… and Each SVR model is trained independently, but using the same input feature vector. This is equivalent to constructing two mapping functions:

[0093]

[0094]

[0095] in, and These are the weight vectors for the temperature and pressure models, respectively. and These are the bias terms for the temperature and pressure models, respectively. It is a kernel function that maps the temperature model input features X to a high-dimensional feature space. It is a kernel function that maps the input features X of the pressure model to a high-dimensional feature space. Although the two models are trained independently, they share the same input feature space, which enables the models to learn from the same set of spectral features. Independent mapping relationships to different physical quantities.

[0096] By establishing regression functions for temperature and pressure respectively, the model can learn from the same set of spectral features. Independent mappings to different physical quantities. During training, each SVR model automatically adjusts its weight vector to maximize the use of features that are most informative for predicting its respective target variable. This design allows the model to intrinsically distinguish the effects of temperature and pressure without requiring explicit physical modeling or step-by-step calculations.

[0097] The real-time demodulation module includes a data acquisition unit, a preprocessing unit, a standardization unit, and a model inference unit.

[0098] Preferably, the real-time demodulation module includes a data acquisition unit, a preprocessing unit, a standardization unit, and a model inference unit, comprising:

[0099] After the sensor starts the measurement, the data acquisition unit immediately acquires the raw spectrum in real time;

[0100] The preprocessing unit extracts feature vectors from the raw spectrum;

[0101] The standardization unit performs standardization processing on the extracted feature vectors to obtain standardized feature vectors.

[0102] The model inference unit inputs the standardized feature vectors into the pre-trained multi-output support vector regression model to obtain the predicted values ​​of temperature and pressure.

[0103] The model output results are checked for reasonableness, including outlier detection. If an outlier is detected, a remeasurement mechanism is triggered.

[0104] In some embodiments, the feature vectors are used in the database constructed in the first step. As input, respectively temperature labels and pressure label As the target, for the two SVR models and Training is then performed. Before training, all input features need to be standardized so that their mean is 0 and their standard deviation is 1, to ensure that features at different scales are equally important to the model.

[0105] The training process involves optimizing key hyperparameters. The following three parameters directly affect the model's generalization ability and prediction accuracy:

[0106] Punishment factor : Control model for exceeding The severity of the penalty for samples from pipelines. A larger value indicates a lower tolerance for training error.

[0107] Insensitive loss parameters Defines the tolerance range for training loss. A larger value indicates that the model is more tolerant of smaller errors.

[0108] Kernel function and its parameters: The preferred kernel function in this invention is a radial basis function, which has the following form: ,in The kernel parameter controls the influence range of a single training sample.

[0109] A strategy combining cross-validation and grid search is employed to optimize these hyperparameters. The calibration dataset is randomly divided into... A mutually exclusive subset (usually) =5 or 10); traverse the hyperparameter combination space defined by the mesh; for each set of hyperparameters, perform... Folded cross-validation—each subset is used sequentially as the validation set, and the remaining subsets are used as the training set. The model is trained on the training set and its predictive performance is evaluated on the validation set; computation... The average performance metric of the cross-validation (usually the root mean square error RMSE) is used as the evaluation metric; finally, the set of hyperparameters that performs best in the cross-validation is selected as the configuration of the final model.

[0110] This step is crucial for ensuring model performance. Through systematic hyperparameter optimization, the final SVR model is optimized. and It exhibits optimal generalization ability, accurately inferring temperature and pressure values ​​from previously unseen spectral features. The use of cross-validation effectively prevents overfitting, ensuring the model's robustness in real-world applications.

[0111] After model training is complete, a comprehensive evaluation of the model's performance is required using an independent test set. Evaluation metrics include:

[0112] Coefficient of determination ( ): Measures the model's ability to explain data variability, calculated as ,in For the sum of squared residuals, This is the total sum of squares.

[0113] Other metrics for evaluating a model include: root mean square error (RMSE). Mean absolute error ( ).

[0114] For temperature model The requirements are: R² ≥ 0.98, RMSE ≤ 0.5℃, MAE ≤ 0.3℃ on the test set; for the pressure model... The requirements are R² ≥ 0.95, RMSE ≤ 0.2%FS, and MAE ≤ 0.15%FS. Only models that meet these performance metrics can be put into practical use.

[0115] This step ensures that the final deployed model meets the accuracy requirements of practical applications. Rigorous performance evaluation not only verifies the model's effectiveness but also provides a reliable quality assurance for subsequent system integration.

[0116] The above is an introduction to the method embodiments. The following describes the solution described in this application through device embodiments.

[0117] Figure 3 This is a block diagram illustrating a temperature and pressure synchronous sensing device for monitoring parameters of a bridge ship collision avoidance system, according to an exemplary embodiment. The device is used in a method for synchronously sensing temperature and pressure parameters of a bridge ship collision avoidance system. (Refer to...) Figure 3 The device includes a sensor calibration module and a real-time demodulation module.

[0118] Sensor calibration module: The sensor calibration module includes a sensor calibration unit, a spectral feature extraction unit, a database unit, a data preprocessing and standardization unit, a model building and training unit, and a hyperparameter optimization unit. The sensor calibration unit is used to collect raw spectral data. The model building and training unit is used to build a multi-output support vector regression model. The construction of the multi-output support vector regression model includes establishing regression functions for temperature and pressure respectively, so that the model learns the independent mapping relationship between temperature and pressure from the same set of spectral features. The multi-output support vector regression model includes a first SVR model and a second SVR model.

[0119] Real-time demodulation module: The real-time demodulation module includes a data acquisition unit, a preprocessing unit, a standardization unit, and a model inference unit.

[0120] A temperature and pressure synchronous sensing device for monitoring parameters of a bridge anti-ship collision system, comprising: a processor; and a memory storing computer-readable instructions, wherein when executed by the processor, the computer-readable instructions implement any one of the above-described methods for synchronous sensing of temperature and pressure for monitoring parameters of a bridge anti-ship collision system.

[0121] A computer-readable storage medium, characterized in that the computer-readable storage medium stores program code, the program code being invoked by a processor to execute the method as described in any one of claims 1 to 7.

[0122] Figure 4This is a schematic diagram of the structure of a temperature and pressure synchronous sensing device for monitoring parameters of a bridge anti-ship collision system provided in an embodiment of the present invention, as shown below. Figure 4 As shown, the temperature and pressure synchronous sensing device for monitoring the parameters of the bridge anti-ship collision system may include the above-mentioned... Figure 3 The illustrated bridge anti-ship collision system parameter monitoring device includes a temperature and pressure synchronous sensing device. Optionally, the temperature and pressure synchronous sensing device 410 for monitoring bridge anti-ship collision system parameters may include a first processor 2001.

[0123] Optionally, the temperature and pressure synchronous sensing device 410 for monitoring the parameters of the bridge anti-ship collision system may also include a memory 2002 and a transceiver 2003.

[0124] The first processor 2001, memory 2002, and transceiver 2003 can be connected via a communication bus.

[0125] The following is combined Figure 4 The following is a detailed description of the various components of the temperature and pressure synchronous sensing device 410 for monitoring parameters of the bridge anti-ship collision system:

[0126] The first processor 2001 is the control center of the temperature and pressure synchronous sensing device 410 for monitoring parameters of the bridge anti-ship collision system. It can be a single processor or a collective term for multiple processing elements. For example, the first processor 2001 can be one or more central processing units (CPUs), application-specific integrated circuits (ASICs), or one or more integrated circuits configured to implement embodiments of the present invention, such as one or more digital signal processors (DSPs), or one or more field-programmable gate arrays (FPGAs).

[0127] Optionally, the first processor 2001 can perform various functions of the temperature and pressure synchronous sensing device 410 for monitoring bridge anti-ship collision system parameters by running or executing software programs stored in the memory 2002 and calling data stored in the memory 2002.

[0128] In a specific implementation, as one example, the first processor 2001 may include one or more CPUs, for example... Figure 4 CPU0 and CPU1 are shown in the diagram.

[0129] In a specific implementation, as one example, the temperature and pressure synchronous sensing device 410 for monitoring the parameters of the bridge anti-ship collision system may also include multiple processors, for example... Figure 4 The first processor 2001 and the second processor 2004 are shown in the diagram. Each of these processors can be a single-core processor or a multi-core processor. Here, a processor can refer to one or more devices, circuits, and / or processing cores used to process data (such as computer program instructions).

[0130] The memory 2002 is used to store the software program that executes the present invention, and is controlled by the first processor 2001 to execute it. The specific implementation method can be referred to the above method embodiment, and will not be repeated here.

[0131] Optionally, the memory 2002 may be a read-only memory (ROM) or other type of static storage device capable of storing static information and instructions, random access memory (RAM) or other type of dynamic storage device capable of storing information and instructions, or electrically erasable programmable read-only memory (EEPROM), compact disc read-only memory (CD-ROM) or other optical disc storage, optical disc storage (including compressed optical discs, laser discs, optical discs, digital universal optical discs, Blu-ray discs, etc.), magnetic disk storage media or other magnetic storage devices, or any other medium capable of carrying or storing desired program code in the form of instructions or data structures and accessible by a computer, but not limited thereto. The memory 2002 may be integrated with the first processor 2001 or may exist independently, and may be connected to the interface circuit of the temperature and pressure synchronous sensing device 410 for monitoring bridge collision prevention system parameters. Figure 4 (Not shown in the image) is coupled to the first processor 2001, and this embodiment of the invention does not specifically limit this.

[0132] The transceiver 2003 is used to communicate with network devices or with terminal devices.

[0133] Alternatively, transceiver 2003 may include a receiver and a transmitter. Figure 4 (Not shown separately). The receiver is used to implement the receiving function, and the transmitter is used to implement the transmitting function.

[0134] Optionally, the transceiver 2003 can be integrated with the first processor 2001, or it can exist independently, and its interface circuit can be used with the temperature and pressure synchronous sensing device 410 for monitoring parameters of the bridge anti-ship collision system. Figure 4 (Not shown in the image) is coupled to the first processor 2001, and this embodiment of the invention does not specifically limit this.

[0135] It should be noted that, Figure 4 The structure of the temperature and pressure synchronous sensing device 410 for monitoring parameters of the bridge anti-ship collision system shown in the figure does not constitute a limitation on the router. The actual knowledge structure identification device may include more or fewer components than shown, or combine certain components, or have different component arrangements.

[0136] Furthermore, the technical effect of the temperature and pressure synchronous sensing device 410 for monitoring the parameters of the bridge anti-ship collision system can be referred to the technical effect of the temperature and pressure synchronous sensing method for monitoring the parameters of the bridge anti-ship collision system described in the above method embodiments, and will not be repeated here.

[0137] It should be understood that the first processor 2001 in the embodiments of the present invention may be a central processing unit (CPU), or it may be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor, etc.

[0138] It should also be understood that the memory in the embodiments of the present invention can be volatile memory or non-volatile memory, or may include both volatile and non-volatile memory. The non-volatile memory can be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory. The volatile memory can be random access memory (RAM), which is used as an external cache. By way of example, but not limitation, many forms of random access memory (RAM) are available, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate synchronous DRAM (DDR SDRAM), enhanced synchronous DRAM (ESDRAM), synchronous linked DRAM (SLDRAM), and direct rambus RAM (DR RAM).

[0139] The above embodiments can be implemented, in whole or in part, by software, hardware (such as circuits), firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, all or part of the processes or functions described in the embodiments of the present invention are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more sets of available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium. A semiconductor medium can be a solid-state drive.

[0140] It should be understood that the term "and / or" in this article is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. A and B can be singular or plural. Additionally, the character " / " in this article generally indicates an "or" relationship between the preceding and following related objects, but it can also represent an "and / or" relationship. Please refer to the context for a more accurate understanding.

[0141] In this invention, "at least one" means one or more, and "more than one" means two or more. "At least one of the following" or similar expressions refer to any combination of these items, including any combination of a single item or a plurality of items. For example, at least one of a, b, or c can represent: a, b, c, ab, ac, bc, or abc, where a, b, and c can be a single item or multiple items.

[0142] It should be understood that, in various embodiments of the present invention, the order of the above-mentioned process numbers does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0143] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0144] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the devices, apparatuses, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0145] In the several embodiments provided by this invention, it should be understood that the disclosed devices, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another device, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0146] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0147] In addition, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0148] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0149] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for synchronously sensing temperature and pressure to monitor parameters of a bridge anti-ship collision system, characterized in that, A system for synchronously sensing temperature and pressure for monitoring parameters in a bridge anti-ship collision system, comprising: a sensor calibration module and a real-time demodulation module, characterized in that it includes: The sensor calibration module includes a sensor calibration unit, a spectral feature extraction unit, a database unit, a data preprocessing and standardization unit, a model building and training unit, and a hyperparameter optimization unit. The sensor calibration unit is used to collect raw spectral data. The model building and training unit is used to build a multi-output support vector regression model. The construction of the multi-output support vector regression model includes establishing regression functions for temperature and pressure respectively, so that the model learns the independent mapping relationship between temperature and pressure from the same set of spectral features. The multi-output support vector regression model includes a first SVR model and a second SVR model. The real-time demodulation module includes a data acquisition unit, a preprocessing unit, a standardization unit, and a model inference unit.

2. The method for synchronously sensing temperature and pressure for monitoring parameters of a bridge anti-ship collision system according to claim 1, characterized in that, The sensor calibration unit includes: The prepared single-mode hollow-coreless fiber optic sensor was placed in a high-precision temperature and pressure control chamber. A sensor calibration task instruction is generated to perform the calibration process and obtain raw spectral data. The sensor calibration task instruction includes defining measurement conditions based on the expected application range, setting calibration points, and for each combination of temperature and pressure values, collecting complete spectral data multiple times using a spectrometer with a spectral resolution of not less than 0.01 nm. The multiple collection of complete spectral data includes collecting multiple samples at each calibration point to reduce the influence of random noise. The calibration points correspond to a series of temperature and pressure value combinations generated within the expected application range.

3. The method for synchronously sensing temperature and pressure for monitoring parameters of a bridge anti-ship collision system according to claim 1, characterized in that, The spectral feature extraction unit includes: Based on the original spectral data, the main interference valley wavelength is extracted from each original spectrum. The extraction of the main interference valley wavelength from each original spectrum includes: accurately locating it using cubic spline interpolation. The main interference valley wavelength corresponds to the wavelength value corresponding to the lowest transmittance in the spectrum. Based on the original spectral data, the wavelengths of secondary interference valleys / peaks are extracted from each original spectrum. The extraction of the wavelengths of secondary interference valleys / peaks from each original spectrum includes: obtaining them by finding local extreme points. The wavelength of the secondary interference valley / peak is the wavelength value of another significant interference extreme point, which is a valley or a peak. Based on the original spectral data, the depth of the main interference valley is calculated from each original spectrum. The calculation of the depth of the main interference valley from each original spectrum includes: calculating based on the light intensity at the bottom of the valley. The depth of the main interference valley is the normalized light intensity value at the bottom of the main interference valley. Based on the original spectral data, the second-order central moment of the spectrum is calculated from each original spectrum. The calculation of the second-order central moment of the spectrum from each original spectrum includes: calculation based on the wavelength of the spectral centroid and the light intensity of the corresponding wavelength. The second-order central moment of the spectrum characterizes the broadening of the spectral energy distribution. Based on the original spectral data, the spectral asymmetry is calculated from each original spectrum. The calculation of the spectral asymmetry from each original spectrum includes: calculation based on the spectral centroid wavelength, the light intensity at the corresponding wavelength, and the second central moment. The spectral asymmetry is used to quantify the asymmetry of the spectral distribution. A multidimensional feature vector is formed based on the wavelength of the primary interference valley, the wavelength of the secondary interference valley / peak, the depth of the primary interference valley, the second-order central moment of the spectrum, and the spectral asymmetry.

4. The method for synchronously sensing temperature and pressure for monitoring parameters of a bridge anti-ship collision system according to claim 1, characterized in that, The database unit includes: At each calibration point, the corresponding spectral feature vector is extracted based on the spectral feature extraction unit; Synchronously record the corresponding real physical quantity labels; Each calibration point, its corresponding feature vector, and its corresponding physical quantity label are assembled into a sample group and added to the database. The physical quantity labels include temperature labels and pressure labels. The database is then subjected to quality checks and annotation improvements to obtain a complete database.

5. The method for synchronously sensing temperature and pressure for monitoring parameters of a bridge anti-ship collision system according to claim 1, characterized in that, The model building and training unit includes: A nonlinear mapping is used to map the input spectral feature vector to a high-dimensional feature space; Two optimal linear regression hyperplanes for temperature and pressure are constructed in this space respectively; Using the same input feature vector, two mapping functions are constructed to train independent SVR models for temperature and pressure, respectively corresponding to the first SVR model and the second SVR model, resulting in a multi-output support vector regression model. The first SVR model includes a mapping function for temperature prediction, which includes the weight vector and bias term of the temperature model. The second regression model includes a mapping function for pressure prediction, which includes the weight vector and bias term of the pressure model.

6. The method for synchronously sensing temperature and pressure for monitoring parameters of a bridge anti-ship collision system according to claim 1, characterized in that, The hyperparameter optimization unit includes: Based on the complete database, extract spectral feature vectors; The spectral feature vectors are transformed into standardized input features using a standardization process. Using standardized input features as input, and temperature and pressure labels as targets, the first and second SVR models are trained respectively. A strategy combining cross-validation and grid search is used to optimize the hyperparameters of the first SVR model and the second SVR model. The cross-validation strategy involves randomly dividing the calibration dataset into mutually exclusive subsets, then using the first subset of each mutually exclusive subset as the validation set and the first subset as the training set, training the model on the training set, and evaluating its prediction performance on the validation set. The grid search strategy involves traversing the hyperparameter combination space defined by the grid. After model training is completed, the performance of the first SVR model and the second SVR model is comprehensively evaluated using an independent test set and preset evaluation metrics.

7. The method for synchronously sensing temperature and pressure for monitoring parameters of a bridge anti-ship collision system according to claim 1, characterized in that, The real-time demodulation module includes a data acquisition unit, a preprocessing unit, a standardization unit, and a model inference unit, comprising: After the sensor starts the measurement, the data acquisition unit immediately acquires the raw spectrum in real time; The preprocessing unit extracts feature vectors from the raw spectrum; The standardization unit performs standardization processing on the extracted feature vectors to obtain standardized feature vectors. The model inference unit inputs the standardized feature vectors into the pre-trained multi-output support vector regression model to obtain the predicted values ​​of temperature and pressure. The model output results are checked for reasonableness, including outlier detection. If an outlier is detected, a remeasurement mechanism is triggered.

8. A temperature and pressure synchronous sensing device for monitoring parameters of a bridge anti-ship collision system, wherein the temperature and pressure synchronous sensing device is used to implement the temperature and pressure synchronous sensing method for monitoring parameters of a bridge anti-ship collision system as described in any one of claims 1-7, characterized in that, The device includes: The sensor calibration module specifically includes a sensor calibration unit, a spectral feature extraction unit, a database unit, a data preprocessing and standardization unit, a model building and training unit, and a hyperparameter optimization unit. The sensor calibration unit is used to collect raw spectral data. The model building and training unit is used to build a multi-output support vector regression model. The construction of the multi-output support vector regression model includes establishing regression functions for temperature and pressure respectively, so that the model learns the independent mapping relationship between temperature and pressure from the same set of spectral features. The multi-output support vector regression model includes a first SVR model and a second SVR model. Real-time demodulation module: Specifically includes a data acquisition unit, a preprocessing unit, a standardization unit, and a model inference unit.

9. A temperature and pressure synchronous sensing device for monitoring parameters of a bridge anti-ship collision system, characterized in that, The bridge anti-ship collision system parameter monitoring temperature and pressure synchronous sensing processor; the memory, the memory storing computer-readable instructions, when executed by the processor, implements the method as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium contains program code that can be invoked by a processor to execute the method as described in any one of claims 1 to 7.

Images