A ship ice structure response cross-region measurement and ice load intelligent identification method and system, and a storage medium

Abstract

The present application belongs to the technical field of load monitoring and identification, and the method comprises two parts of cross-region measurement of ice-induced structural response and intelligent identification of ice load. The cross-region measurement of ice-induced structural response aims to reasonably determine the cross-region measurement mode of ice-induced structural response and implement the measurement based on the spatial distribution of structural strain on the load-bearing member outside the ice load action area. The intelligent identification of ice load aims to build an intelligent identification model that can represent the cross-region mapping mechanism between ice-induced structural response and ice load through a radial basis function neural network, and then input the measured ice-induced structural response into the identification model to obtain accurate and stable ice load identification results. The present application can fully guarantee the accuracy and stability of the ice load identification results in the monitoring scene where the sensor placement area is separated from the ice load action area, has obvious advantages compared with the existing influence coefficient matrix method, and has certain research and application value in the technical field of load monitoring and identification.

Description

Technical Field

[0001] This invention belongs to the field of load monitoring and identification technology, specifically relating to a method, system and storage medium for cross-regional measurement of ship ice-induced structural response and intelligent identification of ice load. Background Technology

[0002] Ice load is an extreme environmental load that threatens the navigation safety of polar vessels. On-ship monitoring is a relatively direct and effective way to obtain ice load information. The technical approach involves deploying strain sensors on load-bearing components such as ribs in typical icebreaking areas like the bow and shoulder, and then inputting the measured ice-induced structural response into an identification model using the influence coefficient matrix method to invert the ice load. However, in actual sensor deployment, limitations such as limited internal space and watertight structures mean that sensors cannot be deployed within the ice load area, necessitating separation of the sensor deployment area from the ice load area. Existing influence coefficient matrix methods struggle to obtain accurate and stable ice load identification results in these scenarios, which is detrimental to ensuring the safety of polar vessels navigating in ice-covered areas. Summary of the Invention

[0003] The purpose of this invention is to provide a method, system and storage medium for cross-regional measurement of ship ice-induced structural response and intelligent identification of ice load, thereby overcoming the technical deficiencies in existing ice load measurement and identification methods.

[0004] The objective of this invention is achieved through the following technical solution:

[0005] A method for cross-regional measurement of ship ice-induced structural response and intelligent identification of ice loads, comprising the following steps:

[0006] Step 1, cross-regional measurement of ice-induced structural response: Based on the simulation model of typical icebreaking parts of the ship, the measurement area and measurement location are determined outside the ice load area, and sensors are installed at the measurement locations to obtain measured data of ice-induced structural response;

[0007] Step 2: Based on the simulation model, construct a dataset containing multiple loading conditions through numerical simulation. The dataset includes ice-induced structural response data samples and corresponding ice load data samples, and divide the dataset into training set and test set according to the proportion.

[0008] Step 3: Construct an intelligent ice load identification model based on a radial basis function neural network, and train the intelligent identification model using a training set. The model is used to establish the mapping relationship between ice-induced structural response and ice load.

[0009] Step 4: Test the trained intelligent recognition model using the test set. When the test results meet the preset requirements, the model training is deemed successful.

[0010] Step 5: Input the measured ice-induced structural response data obtained in Step 1 into the qualified intelligent recognition model, and output the recognition result of the measured ice load.

[0011] Furthermore, the simulation model in step 1 is a full-size finite element model of a typical icebreaking section of the ship under test; the specific steps include:

[0012] Step 1.1: Set up multiple candidate areas for cross-regional measurement of ice-induced structural response outside the ice load area, and set up multiple strain measurement points on the rib webs of each candidate area at half the height of the outer plate, forming a measurement point matrix; divide the ice load area into several sub-regions, and set up various loading conditions according to different load heights, and analyze various strains at each measurement point through finite element analysis; based on the spatial distribution cloud map of various strains in each candidate area, determine the candidate areas with high overall stress level and low amplitude decay rate, as well as the strain type, as the measurement area and observation of ice-induced structural response;

[0013] Step 1.2: Set multiple strain measurement points along the height direction on the rib web within the measurement area as alternative measurement locations; calculate the ice-induced structural response observations at each alternative strain measurement point under different ice load heights using finite element analysis; based on the spatial distribution curve of the alternative observed strains on the rib web, select alternative measurement points with large amplitudes and high amplitude differentiation between adjacent measurement points as the measurement locations for the ice-induced structural response.

[0014] Step 1.3: Delineate the measurement area on the ship to be tested and install strain sensors at the measurement locations to obtain measured data of the ice-induced structural response.

[0015] Furthermore, the candidate areas are set in the vertical, longitudinal, and oblique directions of the ice load area; the various strains include normal strain parallel to the outer plate direction caused by bending deformation, normal strain perpendicular to the outer plate direction caused by extrusion deformation, and shear strain perpendicular to the outer plate direction; the candidate positions are set on the rib webs of the measurement area, at a distance from the outer plate. , , , , , , At the height of the web.

[0016] Furthermore, the loading conditions in step 2 include single sub-region loading, multiple sub-region linear loading, and multiple sub-region random loading.

[0017] Furthermore, the intelligent ice load identification model based on radial basis function neural network in step 3 is as follows:

[0018]

[0019] in, For the first input into the network Group of ice-induced structural response data samples; For input After that, the network's first Ice load identification results output by each output layer neuron; The threshold value in the hidden layer; for With the The weights between neurons in the output layer; For the first The hidden layer neurons and the first The weights between neurons in the output layer; These are the expansion coefficients of the radial basis functions; It is the center of the radial basis functions; This represents the number of neurons in the hidden layer.

[0020] Furthermore, training the intelligent recognition model using the training set includes the following steps:

[0021] Ice-induced structural response data samples in the training set Inject random noise,

[0022]

[0023] in, This is a sample of ice-induced structural response data after noise injection; The noise level is defined as 3% to 5%. Standard deviation; For in the interval Internally generated random numbers;

[0024] The ice load data samples and the ice-induced structural response data samples after noise injection were normalized and converted into interval data. Dimensionless data samples within;

[0025]

[0026]

[0027] in, and These are dimensionless samples of ice-induced structural response and ice load data, respectively. and These are dimensional samples of ice-induced structural response and ice load data, respectively. and These are the minimum and maximum values ​​in the dimensional ice-induced structural response data samples, respectively. and These are the minimum and maximum values ​​in the dimensional ice load data sample, respectively;

[0028] The centers of the radial basis functions are determined using the k-means clustering algorithm; according to the formula... Determine the expansion coefficients of the radial basis functions ,in, The maximum Euclidean norm between the centers of the radial basis functions is used; the weights between the hidden and output layers are determined by solving the pseudo-inverse. ,in, ; ,in For the first The input vector at the th ... The output of each hidden layer neuron; ,in For the first The input vector at the th ... The expected output of each output layer neuron; It is a false rebellion.

[0029] Furthermore, determining the center of the radial basis function includes the following steps:

[0030] (1) Randomly select from the training set Group, The ice-cold structural response data samples were used as the initial cluster centers to determine the number of cluster centers. , ;

[0031] (2) Then randomly select a set of ice-induced structural response data samples from the training set. As input;

[0032] (3) Calculation With the After the nth iteration Cluster centers The Euclidean norm between, and The cluster center with the smallest Euclidean norm belongs to the following category:

[0033]

[0034] (4) Update cluster centers:

[0035]

[0036] in, To learn step length, Only one cluster center is updated at a time, while the other cluster centers remain unchanged.

[0037] (5) Determine whether the algorithm has converged. If the change in cluster centers is less than the preset threshold, then... If the iteration stops, then stop.

[0038]

[0039] Otherwise, return to step (2) to continue iterating; the cluster centers obtained after the iteration is completed are the centers of the radial basis functions.

[0040] Furthermore, the test results in step 4 need to meet preset requirements, including: the true value of ice load in the test set. With identification value Correlation coefficient between And relative error ,

[0041]

[0042]

[0043] in, The sample size of the test set; The standard deviation of the true ice load; The standard deviation of the ice load identification values; The covariance between the true and identified ice load values;

[0044] like and If the requirements are not met, the sample size of the training set is increased, and the recognition model is trained again.

[0045] A computer system includes a memory, a processor, and a computer program stored in the memory, wherein the processor executes the computer program to implement the steps of a method for cross-regional measurement of ship ice-induced structural response and intelligent identification of ice load.

[0046] A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of a method for cross-regional measurement of ship ice-induced structural response and intelligent identification of ice loads.

[0047] The beneficial effects of this invention are as follows:

[0048] (1) Based on the spatial distribution of structural strain on the load-bearing components outside the ice load area, this invention realizes the reasonable determination of the cross-regional measurement method of ice-induced structural response. It can provide reliable structural response input for the ice load identification model in monitoring scenarios where the sensor deployment area is separated from the ice load area.

[0049] (2) This invention constructs an intelligent ice load identification model through a radial basis function neural network, which realizes the accurate characterization of the cross-regional mapping mechanism between ice-induced structural response and ice load. At the same time, it combines noise injection learning algorithm and normalization to improve the fault tolerance and convergence speed of the identification model. In monitoring scenarios where the sensor deployment area and the ice load action area are separated, the accuracy and stability of the ice load identification results can be fully guaranteed.

[0050] (3) The present invention is novel and has strong potential for future expansion. It can be used as a new, efficient and reliable tool for measuring and identifying ice loads on ship structures, and has certain research and application value in the field of load monitoring and identification technology. Attached Figure Description

[0051] Figure 1 This is a schematic diagram of the components of the cross-regional measurement and intelligent ice load identification method for ship ice-induced structural response described in this invention;

[0052] Figure 2 This is a schematic diagram showing the relative positions between the ice load application area and the candidate area for cross-regional measurement of ice-induced structural response as described in this invention.

[0053] Figure 3 This is a schematic diagram showing the arrangement of strain measurement points within the candidate area for cross-regional measurement of ice-induced structural response as described in this invention.

[0054] Figure 4 This is a schematic diagram illustrating the division of the ice load action area and the loading conditions described in this invention;

[0055] Figure 5 This is a schematic diagram of the bending normal strain, compressive normal strain, and shear strain on the rib web of the present invention;

[0056] Figure 6 This is a schematic diagram of the alternative measurement locations within the ice-induced structural response measurement area described in this invention;

[0057] Figure 7 This is a schematic diagram of the radial basis function neural network structure of the ice load intelligent identification model described in this invention;

[0058] Figure 8 These are the test results of the intelligent ice load recognition model described in this invention;

[0059] Figure 9 The results of ice load inversion obtained by the influence coefficient matrix method are shown.

[0060] In the figure: 1 Outer plate; 2 Longitudinal girder; 3 Ribs; 4 Ice load area; 5 Alternative area for cross-regional measurement of ice-induced structural response; 6 Rib web; 7 Strain measurement point; 8 Bending normal strain; 9 Compression normal strain; 10 Shear strain; 11 Alternative measurement location. Detailed Implementation

[0061] The present invention will now be further described with reference to the accompanying drawings.

[0062] The following is combined with Figures 1-7 The specific embodiments of the present invention will be further described below.

[0063] like Figure 1 As shown, this invention provides a method for cross-regional measurement of ship ice-induced structural response and intelligent identification of ice loads, comprising two parts: cross-regional measurement of ice-induced structural response and intelligent identification of ice loads. The specific implementation is as follows:

[0064] Step 1: Cross-regional measurement of ice-induced structural response

[0065] (1-1) Determine the measurement area and observations of the ice-induced structural response.

[0066] First, a full-size finite element model of a typical icebreaking section of the ship under test is established. Figure 2 The model consists of components such as outer plate 1, longitudinal girder 2, and ribs 3. An ice load application area 4 is defined on the model. Alternate regions 5 for measuring the ice-induced structural response across these regions are set vertically, longitudinally, and obliquely along this area. A distance of half the web height from the outer plate 1 is placed on the web of the ribs 6 within each alternate region. A series of strain measurement points are set at point 7, forming a 7x6 matrix of measurement points, as shown below. Figure 3 As shown.

[0067] Then, the ice load application area 4 was divided into 42 sub-areas (7 rows and 6 columns). Based on different load application heights, three loading conditions were set, such as... Figure 4 As shown. The bending normal strain 8, compressive normal strain 9, and shear strain 10 at each measuring point under various loading conditions were calculated using finite element analysis. Figure 5 As shown.

[0068] Finally, based on the spatial distribution cloud maps of the three strains in the vertical, longitudinal, and oblique candidate areas, one candidate area with a higher overall stress level and a lower amplitude decay rate, and one strain type were identified as the measurement area and observation area for the ice-induced structural response.

[0069] (1-2) Determine the measurement location of the ice-induced structural response.

[0070] First, within the defined ice-induced structural response measurement area, the distance from the outer plate to the rib web. , , , , , , ( A series of strain measurement points are set at the web height as alternative measurement locations 11, such as... Figure 6 As shown; then, according to Figure 4 Under the loading conditions described in the experiment, ice loads of different heights were applied within the ice load application area 4. The ice-induced structural response observations at each measuring point were calculated using finite element analysis. Finally, based on the spatial distribution curve of these observations on the rib web, the candidate locations with larger amplitudes and higher amplitude differentiation between measuring points were determined as the measurement locations for the ice-induced structural response.

[0071] (1-3) Conduct cross-regional measurements of ice-induced structural response

[0072] Based on the cross-regional measurement method of ice-induced structural response determined in steps (1-1) and (1-2), the measurement area is delineated on the ship under test, and strain sensors are installed at the measurement locations to obtain the measured data of the ice-induced structural response.

[0073] Step 2: Intelligent Identification of Ice Load

[0074] (2-1) Constructing the training and test sets

[0075] Based on a full-scale finite element model of a typical icebreaking section of the vessel under test, a dataset containing at least 100 loading conditions was constructed through numerical simulation. 80% of the samples were used as the training set, and 20% as the test set. The loading conditions covered single-sub-region loading, linear loading of multiple sub-regions, and random loading of multiple sub-regions. The dataset consisted of two parts: ice-induced structural response data samples and ice load data samples. The former comprised the strain at each measuring point within the ice-induced structural response measurement area, while the latter comprised the ice load in each sub-region within the ice load application area.

[0076] (2-2) Constructing an intelligent identification model for ice load

[0077] Through such Figure 7 The radial basis function neural network shown is used to construct an intelligent identification model for ice load. First, according to the formula... Establish samples of ice-induced structural response data input into the network. ( , (Sample size of the training set) and ice load recognition results of output layer neurons ( , The correspondence between the number of neurons in the output layer and the number of neurons in the output layer.

[0078] (1)

[0079] In the formula, For input The post-network Ice load identification results output by each output layer neuron; The threshold in the hidden layer. ; for With the The weights between neurons in the output layer; This represents the number of neurons in the hidden layer. For the first The hidden layer neurons and the first The weights between neurons in the output layer; For the first The output of each hidden layer neuron, among which ( ) is the center of the radial basis functions.

[0080] Then, according to the formula The formula In Represented as a Gaussian function;

[0081] (2)

[0082] In the formula, These are the expansion coefficients of the radial basis functions.

[0083] Finally, an intelligent recognition model was obtained that can characterize the cross-regional mapping mechanism between ice-induced structural response and ice load.

[0084] (3)

[0085] Ultimately, the ice load identification result output by the entire network is: .

[0086] (2-3) Training an intelligent recognition model for ice load

[0087] First, a noise injection learning algorithm is introduced to improve the fault tolerance of the recognition model, according to the formula... noise level Random noise is injected into the ice-cold structural response data samples in the training set. ,

[0088] (4)

[0089] In the formula, This is a sample of ice-induced structural response data after noise injection; Standard deviation; For in the interval Random numbers generated internally.

[0090] Then, according to the formula Japanese style The ice load data samples and the ice-induced structural response data samples after noise injection were normalized and converted into interval data. Dimensionless data samples within;

[0091] (5)

[0092] (6)

[0093] In the formula, and These are dimensionless samples of ice-induced structural response and ice load data, respectively. and These are dimensional samples of ice-induced structural response and ice load data, respectively. and These are the minimum and maximum values ​​in the dimensional ice-induced structural response data samples, respectively. and These are the minimum and maximum values ​​in the dimensional ice load data sample, respectively;

[0094] Finally, the center of the radial basis function, the spread coefficients, and the weights between the hidden layer and the output layer are determined sequentially, as follows:

[0095] First, the centers of the radial basis functions are determined using the k-means clustering algorithm, following these steps:

[0096] (1) Randomly select from the training set Group( (The number of cluster centers) Ice-cold structural response data samples were used as the initial cluster centers. ( );

[0097] (2) Then randomly select a set of ice-induced structural response data samples from the training set. As input;

[0098] (3) Calculate according to formula (7) With the After the nth iteration Cluster centers The Euclidean norm between, and The cluster center with the smallest Euclidean norm belongs to the following category:

[0099] (7)

[0100] (4) According to the formula Update cluster centers by updating only one cluster center at a time, while keeping the other cluster centers unchanged;

[0101] (8)

[0102] In the formula, To learn step length, Only one cluster center is updated at a time, while the other cluster centers remain unchanged.

[0103] (5) According to the formula To determine if the algorithm has converged, check if the change in cluster centers is less than a pre-set threshold. If the result is positive, stop the iteration; otherwise, proceed to step (2) to continue the iteration; the cluster centers obtained after the iteration is completed are the centers of the radial basis functions.

[0104] (9)

[0105] Then, according to the formula Determine the expansion coefficients of the radial basis functions ,

[0106] (10)

[0107] In the formula, is the maximum Euclidean norm between the centers of the radial basis functions;

[0108] Finally, according to the formula Determine the weights between the hidden layer and the output layer ,

[0109] (11)

[0110] In the formula, ; ,in For the first The input vector at the th ... The output of each hidden layer neuron; ,in For the first The input vector at the th ... The expected output of each output layer neuron; It is a false rebellion.

[0111] (2-4) Intelligent recognition model for testing ice load

[0112] According to the formula Japanese style Calculate the correlation coefficient between the true and identified ice load values ​​in the test set. and relative error ;

[0113] (12)

[0114] (13)

[0115] In the formula, The sample size of the test set; The standard deviation of the true ice load; The standard deviation of the ice load identification values; The covariance between the true and identified ice load values;

[0116] when and If the training set sample size is deemed sufficient and the training effect of the recognition model is deemed satisfactory, the model can be applied to the inversion recognition of measured ice load; otherwise, the training set sample size is increased and the recognition model is trained further.

[0117] (2-5) Applying intelligent recognition models to invert measured ice loads

[0118] Input the cross-regional measurement data of ice-induced structural response obtained in steps (1-3) into the intelligent recognition model, and output the recognition results of the measured ice load.

[0119] Example:

[0120] The advantages of the method of the present invention compared with the existing influence coefficient matrix method are illustrated below through examples, in conjunction with the above-described implementation methods.

[0121] For example Figure 2 The full-size finite element model of a typical icebreaking section of the ship under test, shown, is used to determine the measurement area of ​​the ice-induced structural response as region 5, which is set vertically along the ice load region 4, by implementing step 1. The observations of the ice-induced structural response are then defined as follows: Figure 5 The compressive strain 9 in the figure determined the measurement location of the ice-induced structural response as follows: Figure 6 Position 11 is located on the midrib ventral plate at a distance of 1 / 8 of the ventral plate height from the outer plate. By implementing step 2, training and testing sets are constructed to train and test the intelligent ice load recognition model. The test results are as follows... Figure 8 As shown, most of the output points (ice load identification values) fall on the target line (true ice load values), and the correlation coefficient between the ice load identification values ​​and the true values ​​is... And relative error The training and testing of the intelligent recognition model are completed, meeting the preset requirements.

[0122] The existing influence coefficient matrix method was used to invert the ice load in the above test set. The inversion results are as follows: Figure 9 As shown, the correlation coefficient between the ice load identification value obtained by the influence coefficient matrix method and the true value is... relative error Compared with the influence coefficient matrix method, the inversion results of the method of the present invention show that more output points (ice load identification values) fall on the target line (ice load true values), and the correlation coefficient between the ice load identification values ​​and the true values ​​is larger and the relative error is smaller. This indicates that the method of the present invention has higher ice load inversion accuracy and stability in monitoring scenarios where the sensor deployment area and the ice load action area are separated.

[0123] In particular, in some preferred embodiments of the present invention, a computer device is also provided, including a memory and a processor and a computer program stored in the memory, wherein the processor executes the computer program to implement the steps of the cross-regional measurement and intelligent identification method for ship ice-induced structural response described in any of the above embodiments.

[0124] In some other preferred embodiments of the present invention, a computer-readable storage medium is also provided, on which a computer program / instruction is stored, wherein when the computer program is executed by a processor, the steps of the method for cross-regional measurement of ship ice-induced structural response and intelligent identification of ice load described in any of the above embodiments are implemented.

[0125] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the above embodiments of the method for cross-regional measurement of ship ice-induced structural response and intelligent identification of ice load, which will not be repeated here.

[0126] Computer-readable storage media encompass a variety of types, including persistent and non-persistent, portable and fixed. These media store information using different technologies, and the content can be machine instructions, data structures, program modules, or other types of data. Some typical examples of computer storage media include: phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), various types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory and other storage technologies, optical storage media such as CD-ROM and digital video disc (DVD), magnetic storage devices such as magnetic tape and disks, and other non-transferable media used to store information accessible to computing devices. It is important to note that the computer-readable media described herein do not include temporary storage media, such as modulated data signals and carrier waves.

[0127] Those skilled in the art will further recognize that the operation of the module can be achieved using existing technical protocols or programs, without relying on new computer programs themselves. The units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. 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.

[0128] The steps of the methods or algorithms described in conjunction with the embodiments disclosed herein can be implemented in hardware, a software module executed by a processor, or a combination of both. The software module can be located in random access memory (RAM), main memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disk, removable disk, CD-ROM, or any other form of storage medium known in the art.

[0129] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for cross-regional measurement of ship ice-induced structural response and intelligent identification of ice load, characterized in that, Includes the following steps: Step 1, cross-regional measurement of ice-induced structural response: Based on the simulation model of typical icebreaking parts of the ship, the measurement area and measurement location are determined outside the ice load area, and sensors are installed at the measurement locations to obtain measured data of ice-induced structural response; Step 2: Based on the simulation model, construct a dataset containing multiple loading conditions through numerical simulation. The dataset includes ice-induced structural response data samples and corresponding ice load data samples, and divide the dataset into training set and test set according to the proportion. Step 3: Construct an intelligent ice load identification model based on a radial basis function neural network, and train the intelligent identification model using a training set. The model is used to establish the mapping relationship between ice-induced structural response and ice load. Step 4: Test the trained intelligent recognition model using the test set. When the test results meet the preset requirements, the model training is deemed successful. Step 5: Input the measured ice-induced structural response data obtained in Step 1 into the qualified intelligent recognition model, and output the recognition result of the measured ice load.

2. The method for cross-regional measurement of ship ice-induced structural response and intelligent identification of ice load according to claim 1, characterized in that, The simulation model in step 1 is a full-size finite element model of a typical icebreaking section of the ship under test; the specific steps include: Step 1.1: Set up multiple candidate areas for cross-regional measurement of ice-induced structural response outside the ice load area, and set up multiple strain measurement points on the rib webs of each candidate area at half the height of the outer plate, forming a measurement point matrix; divide the ice load area into several sub-regions, and set up various loading conditions according to different load heights, and analyze various strains at each measurement point through finite element analysis; based on the spatial distribution cloud map of various strains in each candidate area, determine the candidate areas with high overall stress level and low amplitude decay rate, as well as the strain type, as the measurement area and observation of ice-induced structural response; Step 1.2: Set multiple strain measurement points along the height direction on the rib web within the measurement area as alternative measurement locations; calculate the ice-induced structural response observations at each alternative strain measurement point under different ice load heights using finite element analysis; based on the spatial distribution curve of the alternative observed strains on the rib web, select alternative measurement points with large amplitudes and high amplitude differentiation between adjacent measurement points as the measurement locations for the ice-induced structural response. Step 1.3: Delineate the measurement area on the ship to be tested and install strain sensors at the measurement locations to obtain measured data of the ice-induced structural response.

3. The method for cross-regional measurement of ship ice-induced structural response and intelligent identification of ice load according to claim 2, characterized in that, The candidate areas are set in the vertical, longitudinal, and oblique directions of the ice load area; the various strains include normal strain parallel to the outer plate direction caused by bending deformation, normal strain perpendicular to the outer plate direction caused by extrusion deformation, and shear strain perpendicular to the outer plate direction; the candidate positions are set on the rib webs of the measurement area, at a distance from the outer plate. , , , , , , At the height of the web.

4. The method for cross-regional measurement of ship ice-induced structural response and intelligent identification of ice load according to claim 1, characterized in that, The loading conditions in step 2 include single sub-region loading, linear loading of multiple sub-regions, and random loading of multiple sub-regions.

5. The method for cross-regional measurement of ship ice-induced structural response and intelligent identification of ice load according to claim 1, characterized in that, The intelligent ice load identification model based on radial basis function neural network in step 3 is as follows: in, For input The post-network Ice load identification results output by each output layer neuron; For the first input into the network Group of ice-induced structural response data samples; The threshold value in the hidden layer; for With the The weights between neurons in the output layer; For the first The hidden layer neurons and the first The weights between neurons in the output layer; These are the expansion coefficients of the radial basis functions; It is the center of the radial basis functions; This represents the number of neurons in the hidden layer.

6. The method for cross-regional measurement of ship ice-induced structural response and intelligent identification of ice load according to claim 5, characterized in that, The training of the intelligent recognition model using the training set includes the following steps: Ice-induced structural response data samples in the training set Inject random noise, in, This is a sample of ice-induced structural response data after noise injection; The noise level is defined as 3% to 5%. Standard deviation; For in the interval Internally generated random numbers; The ice load data samples and the ice-induced structural response data samples after noise injection were normalized and converted into interval data. Dimensionless data samples within; In the formula, and These are dimensionless samples of ice-induced structural response and ice load data, respectively. and These are dimensional samples of ice-induced structural response and ice load data, respectively. and These are the minimum and maximum values ​​in the dimensional ice-induced structural response data samples, respectively. and These are the minimum and maximum values ​​in the dimensional ice load data sample, respectively; The centers of the radial basis functions are determined using the k-means clustering algorithm; according to the formula... Determine the expansion coefficients of the radial basis functions ,in, The maximum Euclidean norm between the centers of the radial basis functions is used; the weights between the hidden and output layers are determined by solving the pseudo-inverse. ,in, ; ,in For the first The input vector at the th ... The output of each hidden layer neuron; ,in For the first The input vector at the th ... The expected output of each output layer neuron; It is a false rebellion.

7. The method for cross-regional measurement of ship ice-induced structural response and intelligent identification of ice load according to claim 6, characterized in that, Determining the center of the radial basis function includes the following steps: (1) Randomly select from the training set Group, The ice-cold structural response data samples were used as the initial cluster centers to determine the number of cluster centers. , ; (2) Then randomly select a set of ice-induced structural response data samples from the training set. As input; (3) Calculation With the After the nth iteration Cluster centers The Euclidean norm between, and The cluster center with the smallest Euclidean norm belongs to the following category: (4) Update cluster centers: in, To learn step length, Only one cluster center is updated at a time, while the other cluster centers remain unchanged. (5) Determine whether the algorithm has converged. If the change in cluster centers is less than the preset threshold, then... If the iteration stops, then stop. Otherwise, return to step (2) to continue iterating; the cluster centers obtained after the iteration is completed are the centers of the radial basis functions.

8. The method for cross-regional measurement of ship ice-induced structural response and intelligent identification of ice load according to claim 6, characterized in that, The test results in step 4 need to meet preset requirements, including: the true value of ice load in the test set. With identification value Correlation coefficient between And relative error , in, The standard deviation of the true ice load; The standard deviation of the ice load identification values; The covariance between the true and identified ice load values; like and If the requirements are not met, the sample size of the training set is increased, and the recognition model is trained again.

9. A computer system comprising a memory, a processor, and a computer program stored in the memory, characterized in that: The processor executes the computer program to implement the steps of the method according to any one of claims 1 to 8.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When executed by a processor, the computer program implements the steps of the method according to any one of claims 1 to 8.

Images