Steady-state identification method based on multi-element and multi-dimensional characteristics of cold atom absolute gravimeter

Abstract

The invention relates to the related technical field of atomic gravity measurement, and provides a steady-state recognition method based on multi-element and multi-dimensional characteristics of a cold atom absolute gravimeter, and the method comprises the steps: S1, collecting the historical data of the operation of a ship body based on the cold atom absolute gravimeter at each steady-state level, the historical data comprising the acceleration and inclination angle of the ship body, and the acceleration and inclination angle of the ship body; the atomic number data of the cold atom absolute gravimeter are obtained; s2, dividing the acceleration, the inclination angle and the atomic number data into feature groups to form a multi-element multi-dimensional ship body steady-state feature matrix; s3, based on the ship body steady-state characteristic matrix and a machine learning algorithm, establishing a ship body steady-state identification model, and performing cross validation on the ship body steady-state identification model; and S4, according to the ship body steady state identification model after cross validation, identifying the ship body steady state based on the currently acquired acceleration, inclination angle and atomic number data of the ship body. According to the multi-element multi-dimensional information of the shipborne cold atom absolute gravimeter, the ship stability identification precision is further improved.

Description

Technical Field

[0001] The present invention relates to the technical field related to atomic gravity measurement, and more specifically, to a steady-state identification method, system, electronic equipment and storage medium based on multi-factor and multi-dimensional characteristics of a cold atomic absolute gravimeter. Background Art

[0002] The cold atom absolute gravimeter is a high-precision gravity measurement device based on the principle of atomic interference. Compared with traditional relative gravimeters, it has the advantages of no drift, no mechanical wear, and long-term continuous gravity measurement. Generally, this type of equipment has a complex structure, especially the ship-borne cold atom absolute gravimeter. As the ship moves, it will inevitably experience attitude changes such as left and right shaking and up and down sinking. The ship-borne atomic gravimeter generally needs to use its inertial stabilization platform, three-axis acceleration data, inclination data, etc. to dynamically compensate for the measured results of each module. Therefore, the steady-state information of the hull is of great significance to the high-precision measurement of the atomic gravimeter and the safety of ship navigation.

[0003] Nowadays, for all kinds of ships, when sailing at sea, the inclination data at the time of observation (or the roll, pitch and settlement of the attitude meter, etc.) are generally used to characterize the left and right, front and back or up and down vibration of the hull in real time, so as to evaluate the stability of the hull, the external sea conditions and the wind and wave conditions, and use their feedback correction to sail more smoothly and safely, and to survey and measure multiple factors including gravity in the waters. However, the observation time of the inclinometer is often instantaneous, and the attitude changes such as the longitudinal, roll and settlement of the hull are often completed within a specific period, rather than instantaneously. Therefore, using a single and instantaneous inclination data to evaluate the steady state of the hull often has a large error, which causes misjudgment of navigation factors such as the hull attitude, external wind and waves and sea conditions. The ship-borne cold atom absolute gravimeter can not only measure the absolute gravity during the navigation process, but also the real-time data or combination of data of each component module contains more information that can reflect the operating status of its carrier, and the hull steady state information is an important parameter that can be deeply mined.

[0004] Based on this, the present invention aims to provide a solution for further improving the identification of ship hull stability in atomic gravity measurement technology based on multi-factor and multi-dimensional information of a ship-borne cold atomic absolute gravimeter. Summary of the invention

[0005] In view of the shortcomings of existing hull steady-state characterization and identification methods, the present invention provides a hull steady-state identification method based on the multi-factor and multi-dimensional characteristics of a cold atom absolute gravimeter. By extracting multi-factor information of the ship-borne cold atom absolute gravimeter and combining it with a machine learning algorithm, a more accurate judgment of the hull steady state is made.

[0006] According to a first aspect of the present invention, a steady-state recognition method based on multi-element multi-dimensional characteristics of a cold-atom absolute gravimeter is provided, including the following steps: S1. At each steady-state level, historical data of the hull operation is collected based on the cold-atom absolute gravimeter, and the historical data includes the acceleration, inclination angle of the hull, and the atomic number data of the cold-atom absolute gravimeter; S2. The acceleration, inclination angle, and atomic number data are divided into feature groups to form a multi-element multi-dimensional hull steady-state feature matrix; S3. Based on the hull steady-state feature matrix and a machine learning algorithm, a hull steady-state recognition model is established, and cross-validation is performed on the hull steady-state recognition model; S4. According to the hull steady-state recognition model after cross-validation, the hull steady state is recognized based on the currently collected acceleration, inclination angle, and atomic number data of the hull.

[0007] Based on the above technical solution, the present invention can also be improved as follows.

[0008] Optionally, step S1 includes: Collecting the three-axis acceleration data of the X, Y, and Z axes of the hull operation at each steady-state level based on the cold-atom absolute gravimeter; Collecting the inclination angle states of the X and Y axes of the hull operation at each steady-state level based on the cold-atom absolute gravimeter; Collecting the total atomic number and the three-state atomic number of the cold-atom absolute gravimeter of the hull operation at each steady-state level according to the interference and state selection output results of cold atoms.

[0009] Optionally, step S2 includes: S201. For the data collected at the same steady-state level, the acceleration data of each axis, the inclination angle state data of each axis, the total atomic number, and the three-state atomic number are respectively used as a type of single-element feature data, and each type of single-element feature data is divided into multiple feature groups according to a preset dimension to form multiple single-element multi-dimensional steady-state feature matrices; S202. Combining all the single-element multi-dimensional steady-state feature matrices at the same steady-state level to obtain a multi-element multi-dimensional steady-state feature matrix at a single steady-state level; S203. Combining the multi-element multi-dimensional steady-state feature matrices at all steady-state levels to form a hull steady-state feature matrix.

[0010] Optionally, in step S201, the step of dividing each type of single-element feature data into multiple feature groups according to a preset dimension to form multiple single-element multi-dimensional steady-state feature matrices includes: Arranging a certain type of single-element feature data in columns; Divide the single-element feature data of the current category into multiple feature groups according to the preset number of rows or time period; Arrange the multiple divided feature groups in rows or columns to form a single-element multi-dimensional steady-state feature matrix of the current category; Traverse the single-element feature data of all categories to obtain multiple single-element multi-dimensional steady-state feature matrices.

[0011] Optionally, in step S201, the dividing of the single-element feature data of each category into multiple feature groups according to the preset dimension to form multiple single-element multi-dimensional steady-state feature matrices includes: Arrange the single-element feature data of a certain category in rows; Divide the single-element feature data of the current category into multiple feature groups according to the preset number of columns or time period; Arrange the multiple divided feature groups in rows or columns to form a single-element multi-dimensional steady-state feature matrix of the current category; Traverse the single-element feature data of all categories to obtain multiple single-element multi-dimensional steady-state feature matrices.

[0012] Optionally, in step S1, it further includes: collecting hydrological and meteorological feature data of the water area where the ship is located, and the hydrological and meteorological feature data includes one or more of heading and speed, flow direction and velocity, wind direction and speed, and surge condition; In step S201, it further includes: For the hydrological and meteorological feature data collected at the same steady-state level, according to the number of items of the collected hydrological and meteorological feature data, use the hydrological and meteorological feature data as one or more categories of single-element feature data; Divide the single-element feature data of each category into multiple feature groups according to the preset dimension to form a single or multiple single-element multi-dimensional steady-state feature matrices corresponding to the number of items of the hydrological and meteorological feature data.

[0013] Optionally, in step S3, the machine learning algorithm is one of the linear discriminant analysis algorithm, support vector machine algorithm, random forest algorithm, neural network algorithm, and partial least squares algorithm; the cross-validation method used is K-fold cross-validation, leave-one-out cross-validation, or leave-K cross-validation.

[0014] According to the second aspect of the present invention, a steady-state recognition system based on multi-element multi-dimensional features of a cold atom absolute gravimeter is provided, including: An acquisition module, configured to collect historical data of the hull operation based on a cold atom absolute gravimeter at each steady-state level, where the historical data includes the acceleration and inclination of the hull, and the atomic number data of the cold atom absolute gravimeter; A processing module, configured to divide the acceleration, inclination, and atomic number data into feature groups to form a multi-element multi-dimensional hull steady-state feature matrix; A construction module, configured to establish a hull steady-state recognition model based on the hull steady-state feature matrix and a machine learning algorithm, and perform cross-validation on the hull steady-state recognition model; An identification module, configured to identify the hull steady state based on the cross-validated hull steady-state recognition model and the acceleration, inclination, and atomic number data of the currently collected hull.

[0015] According to a third aspect of the present invention, there is provided an electronic device, including a memory and a processor, where the processor is configured to implement the steps of the above-mentioned steady-state recognition method based on multi-element and multi-dimensional features of a cold atom absolute gravimeter when executing a computer management program stored in the memory.

[0016] According to a fourth aspect of the present invention, there is provided a computer-readable storage medium, on which a computer management program is stored, and the computer management program is configured to implement the steps of the above-mentioned steady-state recognition method based on multi-element and multi-dimensional features of a cold atom absolute gravimeter when executed by a processor.

[0017] A steady-state recognition method, system, electronic device, and storage medium based on multi-element and multi-dimensional features of a cold atom absolute gravimeter provided by the present invention can more accurately judge the hull stability state according to the multi-element and multi-dimensional information of an on-board cold atom absolute gravimeter and in combination with a machine learning algorithm. The beneficial effects of the present invention are mainly manifested in: 1. The present invention can use the multi-element features of a cold atom absolute gravimeter to discriminate the hull steady state, which is more accurate than traditional single-element and single-dimensional features (such as a single inclination feature, acceleration feature, or attitude feature at a certain moment, etc.). 2. The present invention divides and discriminates the multi-element features in groups, and performs hull steady-state recognition by instantaneously discriminating data of a specific number or duration in the past. This number or duration far exceeds the hull's swaying and settlement periods, and the judgment result is more continuous and reliable. 3. The present invention first introduces the atomic number of cold atom absolute gravity, which is susceptible to the hull attitude, as a hull steady-state discrimination feature, enhancing the diversity of discrimination features and the accuracy of the result. Description of the Drawings

[0018] Figure 1 It is a schematic flowchart of a hull steady-state recognition method based on multi-element and multi-dimensional features of a cold atom absolute gravimeter provided by the present invention; Figure 2 It is a schematic diagram of the multi-element feature dimension transformation and combination process provided in an embodiment; Figure 3 It is a schematic diagram of a hull steady-state recognition model based on multi-element and multi-dimensional features and a linear discriminant analysis algorithm and its cross-validation provided in an embodiment; Figure 4Comparison diagram of discrimination results of hull steady-state recognition model based on single-element single-dimensional features and multi-element multi-dimensional features provided for a certain embodiment; Figure 5 Block diagram of the composition of a steady-state recognition system based on multi-element multi-dimensional features of a cold atom absolute gravimeter provided by the present invention; Figure 6 Schematic diagram of the hardware structure of a possible electronic device provided by the present invention; Figure 7 Schematic diagram of the hardware structure of a possible computer-readable storage medium provided by the present invention. Specific embodiments

[0019] The following combines the accompanying drawings and embodiments to further describe in detail the specific embodiments of the present invention. The following embodiments are used to illustrate the present invention, but are not used to limit the scope of the present invention.

[0020] Figure 1 Flowchart of a steady-state recognition method based on multi-element multi-dimensional features of a cold atom absolute gravimeter provided by the present invention, as Figure 1 shown, the method includes steps S1 to S4: S1, at each steady-state level, collect historical data of the hull operation based on the cold atom absolute gravimeter, and the historical data includes the acceleration, inclination angle of the hull, and the atomic number data of the cold atom absolute gravimeter; S2, divide the acceleration, inclination angle, and atomic number data into feature groups to form a multi-element multi-dimensional hull steady-state feature matrix; S3, based on the hull steady-state feature matrix and machine learning algorithms, establish a hull steady-state recognition model, and perform cross-validation on the hull steady-state recognition model; S4, based on the hull steady-state recognition model after cross-validation, recognize the hull steady state based on the currently collected acceleration, inclination angle, and atomic number data of the hull.

[0021] It can be understood that, based on the defects in the background technology, the embodiment of the present invention proposes a method for recognizing the hull steady state based on multi-element multi-dimensional features of a cold atom absolute gravimeter. This method combines machine learning algorithms with the multi-element multi-dimensional information of an on-board cold atom absolute gravimeter to make a more accurate judgment on the hull steady state. The beneficial effects of this embodiment are mainly manifested in: 1. This embodiment can use the multi-element features of the cold atom absolute gravimeter to discriminate the hull steady state, which is more accurate than traditional single-element single-dimensional features (such as a single inclination angle feature, acceleration feature, or attitude feature at a certain moment, etc.); 2. In this embodiment, multi-element features are grouped and discriminated, and the steady state of the hull is identified by instantaneously discriminating data of a specific quantity or duration in the past, where the quantity or duration far exceeds the swaying and settlement periods of the hull, and the judgment result is more continuous and reliable. 3. In this embodiment, for the first time, the number of atoms of cold atom absolute gravity that is vulnerable to the influence of the hull attitude is introduced as a discriminant feature for the steady state of the hull, enhancing the diversity of the discriminant features and the accuracy of the result.

[0022] In a possible embodiment, step S1 includes sub-steps S101 to S103: S101, Turn on the on-board cold atom absolute gravimeter and make it in a stable normal working state. Determine the number d of steady state levels, make the hull travel stably at one of the steady state levels, and determine that the cold atom absolute gravimeter works normally on the hull at the current steady state level.

[0023] S102, At the current steady state level, based on the acceleration detection module of the cold atom absolute gravimeter, collect the real-time three-axis acceleration data of the hull in the X, Y, and Z directions at each steady state level, and record them as 、 、 respectively, where m is the total number of rows of the collected data and n is the number of columns of the collected data; Based on the inclination detection module of the cold atom absolute gravimeter, collect the real-time inclination states of the hull in the X and Y directions at each steady state level, and record them as: 、 ; According to the interference and state selection output results of the cold atoms in the cold atom absolute gravimeter, collect the real-time total number of atoms and the number of three-state atoms of the cold atom absolute gravimeter when the hull is running at the current steady state level. The total number of atoms is recorded as , and the number of three-state atoms is recorded as .

[0024] S103, Traverse the d steady state levels to obtain the X, Y, and Z three-axis acceleration data of the hull at each steady state level, as well as the X and Y axis inclination states and the total number of atoms and the number of three-state atoms of the cold atom absolute gravimeter.

[0025] It can be understood that the types of features of the hull obtained in step S1 at least include the three-axis acceleration data / / , the X / Y axis inclination state data / , the total number of atoms recorded as and the number of three-state atoms recorded as . Each type of feature can be used as a type of single-element feature to participate in the subsequent steps.

[0026] As a preferred solution, in order to further improve the accuracy of hull steady state recognition, other hydrological and meteorological characteristic data of the water area where the ship is located, such as heading and speed, flow direction and velocity, wind direction and speed, and surge condition, can also be included in the optional features.

[0027] Therefore, in one possible implementation, step S1 further includes: Collect hydrological and meteorological characteristic data of the water area where the ship is located, and the hydrological and meteorological characteristic data includes one or more of heading and speed, flow direction and velocity, wind direction and speed, and surge condition. Use the hydrological and meteorological characteristic data as element features to participate in the subsequent steps.

[0028] In one possible implementation manner, step S2 includes sub-steps S201 to S203.

[0029] S201, as Figure 2 shown, for various types of characteristic data collected under the same steady state level, each type of characteristic data (such as each-axis acceleration data, each-axis inclination state data, total atomic number, three-state atomic number, heading and speed, flow direction and velocity, wind direction and speed, and surge condition) is respectively used as a type of single-element characteristic data, and the various types of single-element characteristic data are divided into multiple characteristic groups according to a preset dimension to form multiple single-element multi-dimensional steady state characteristic matrices.

[0030] More specifically, in one possible implementation, in step S201, the process of dividing the various types of single-element characteristic data into multiple characteristic groups according to a preset dimension to form multiple single-element multi-dimensional steady state characteristic matrices includes: Arrange a certain type of single-element characteristic data in columns; According to the preset number of rows or time period, divide the single-element characteristic data of the current category into multiple characteristic groups; Arrange the divided multiple characteristic groups in rows or columns to form a single-element multi-dimensional steady state characteristic matrix of the current category; Traverse all types of single-element characteristic data to obtain multiple single-element multi-dimensional steady state characteristic matrices.

[0031] Or, as another implementation manner parallel to the previous embodiment, in step S201, the process of dividing the various types of single-element characteristic data into multiple characteristic groups according to a preset dimension to form multiple single-element multi-dimensional steady state characteristic matrices includes: Arrange a certain type of single-element characteristic data in rows; According to the preset number of columns or time period, divide the single-element characteristic data of the current category into multiple characteristic groups; Arrange the divided multiple characteristic groups in rows or columns to form a single-element multi-dimensional steady state characteristic matrix of the current category; Traverse the single-element feature data of all categories to obtain multiple single-element multi-dimensional steady-state feature matrices.

[0032] To more clearly illustrate the generation process of the single-element multi-dimensional feature matrix in step S2, the following is combined with Figure 2 for an example. Figure 2 shows a schematic diagram of the multi-element feature dimension transformation and combination process at a single steady-state level. Combining Figure 2 , first, an example of the processing process of the 3-axis acceleration data is given. The collected 3-axis acceleration data / / is regarded as three categories of single-element features. The acceleration data of each axis is arranged by column, and then three columns of acceleration data representing three categories of single-element features are obtained. Then, according to the same grouping rule, each column of acceleration data is divided into feature groups with a rows per group. The multiple feature groups formed by dividing the same category of single-element features are arranged row by row or column by column to form a single-element multi-dimensional feature matrix, whose dimension is , where c = m / a, a is the number of data rows in each feature group, m is the total number of data rows of the current single-element feature, and c is the number of feature group columns obtained from the current single-element feature. After processing the three-axis acceleration data by the above method, three single-element multi-dimensional feature matrices corresponding to the acceleration data of the X, Y, and Z axes are obtained. Then, the element features such as the X and Y axis inclination data, the total number of atoms, and the number of three-state atoms collected are also divided into feature groups according to the above steps, and the corresponding single-element multi-dimensional feature matrices are obtained, and their dimensions are all .

[0033] Considering the hydrological and meteorological feature data of the water area where the ship is located, in step S201, it further includes: For the hydrological and meteorological feature data collected at the same steady-state level, according to the number of items of the collected hydrological and meteorological feature data, the hydrological and meteorological feature data is used as one or more categories of single-element feature data; Each category of single-element feature data is divided into multiple feature groups according to a preset dimension to form one or more single-element multi-dimensional steady-state feature matrices corresponding to the number of items of the hydrological and meteorological feature data.

[0034] S202. Combine all the single-element multi-dimensional steady-state feature matrices at the same steady-state level, for example, combine them row by row, to obtain a multi-element multi-dimensional steady-state feature matrix including all the considered element features at a single steady-state level , where k is the type of single-element feature, is the dimension of the single-element multi-dimensional feature matrix; for example, if the number of types k of the single-element features considered is 7, including 3-axis acceleration data, X / Y-axis inclination data, total number of atoms, and number of three-state atoms, then the multi-element multi-dimensional steady-state feature matrix obtained is 。

[0035] S203. Assuming the number of steady-state levels is d, arrange the single-class single-element multi-dimensional feature groups under d steady-state levels side by side, and combine them to form the hull steady-state feature matrix that includes all the element features considered under all steady-state levels 。

[0036] In a possible implementation manner, in step S3, based on the multi-element multi-dimensional hull steady-state feature matrix obtained in step S2 and a machine learning algorithm, establish a hull steady-state recognition model and perform cross-validation on this model.

[0037] Among them, the machine learning algorithm is one of the linear discriminant analysis algorithm, support vector machine algorithm, random forest algorithm, neural network algorithm, and partial least squares algorithm; the cross-validation method adopted is K-fold cross-validation, leave-one-out cross-validation, or leave-K cross-validation.

[0038] Then, use the multi-element multi-dimensional hull steady-state recognition model established in step S3 to predict the multi-element hull steady-state feature data in the reserved test set and output the prediction result.

[0039] Finally, collect the current acceleration, inclination, and atomic number data of the hull. In step S4, input the currently collected acceleration, inclination, and atomic number data of the hull into the hull steady-state recognition model after cross-validation to output the recognition result of the hull steady state.

[0040] Now, in combination with a specific implementation scenario, verify the technical effect of the present invention.

[0041] S1. At each steady-state level, based on the cold atom absolute gravimeter, collect the historical data of the hull operation. The historical data includes the acceleration, inclination of the hull, and the atomic number data of the cold atom absolute gravimeter. The specific operation steps are as follows: First, turn on the on-board cold atom absolute gravimeter and make it in a stable normal working state; Then, use the acceleration detection module of the cold atom absolute gravimeter to collect the acceleration data in the X, Y, and Z axis directions of the hull for 1 hour at a speed of 10 knots under 1 to 4 different stable states (or sea conditions). Since the data sampling rate of this cold atom absolute gravimeter is 0.6 (that is, data is collected once every 0.6 seconds), a total of 6000 data can be collected, which are respectively recorded as: 、 、 ; Meanwhile, by using the inclinometer module of the cold atom absolute gravimeter, the inclination states in two directions of the X and Y axes of the hull are collected. The collection parameters and modes are the same as those in the acceleration data collection step. Therefore, the inclination state data of the X and Y axes can be respectively recorded as: , ; Meanwhile, according to the interference and state selection output results of cold atoms, the total number of atoms and the number of three-state atoms of the cold atom absolute gravimeter are collected. The collection parameters and modes are the same as those in the acceleration data collection step. The obtained total number of atoms and three-state atom number data can be respectively recorded as: , .

[0042] S2. Divide the acceleration, inclination, and atom number data into feature groups to form a multi-factor and multi-dimensional hull steady-state feature matrix. The specific operation steps are as follows: First, in the 3-axis acceleration data collected, the acceleration data of each axis is arranged column by column to obtain 3 columns of acceleration data. Each column of acceleration data is divided into feature groups with a = 10 rows per group, and a total of c = 6000 / 10 = 600 feature groups can be obtained; then all the feature groups are arranged side by side to form a single-factor multi-dimensional feature matrix, and its dimension is . Then, the collected X-axis inclination data, Y-axis inclination data, total number of atoms, and number of three-state atoms are divided into feature groups according to the above steps, and the corresponding single-factor multi-dimensional feature matrices are obtained, and their dimensions are all ; Second, combine all the single-factor multi-dimensional feature matrices obtained in the previous step row by row to obtain a multi-factor multi-dimensional feature matrix at a single steady-state level (that is ), and the process is as Figure 2 shown; Then, it can be known from step S1 that there are d = 4 steady-state levels in this embodiment. Then, all the single-class multi-dimensional feature groups at all steady-state levels are arranged side by side. Therefore, the overall multi-factor multi-dimensional feature matrix at all steady-state levels can be obtained (that is, the multi-factor multi-dimensional hull steady-state feature matrix ).

[0043] S3. Based on the hull steady-state feature matrix and machine learning algorithm, establish a hull steady-state recognition model and perform cross-validation on the hull steady-state recognition model. The specific operation steps are as follows: Take the overall multi-factor multi-dimensional feature matrix obtained in step S2 (that is, the multi-factor multi-dimensional hull steady-state feature matrix It is input into the linear discriminant analysis algorithm to establish a hull steady-state recognition model based on the linear discriminant analysis algorithm, and the model is cross-validated by means of 6-fold cross-validation. The process is as Figure 3 shown. After verification, the accuracy rate of the cross-validation of the hull steady-state recognition model is as Figure 4 shown.

[0044] Figure 4 The hull steady-state recognition accuracy rates considering various single-factor features and multi-factor multi-dimensional features respectively are statistically analyzed. From Figure 4 it can be seen that the accuracy rate of the hull steady-state recognition method considering multi-factor multi-dimensional features in the embodiment of the present invention is far superior to other solutions.

[0045] S4. Collect the current acceleration, inclination angle and atomic number data of the hull, input them into the hull steady-state recognition model after cross-validation, and output the hull steady-state recognition result.

[0046] In this step, the hull steady-state recognition model established in step S3 and cross-validated is used to predict 100 groups of test data with multi-factor multi-dimensional features in the test set, and the prediction results are output, as shown in Table 1. In Table 1, if the predicted level is the same as the standard level, it is a correct prediction, and the dark gray box is the label of the wrong prediction. From the results in Table 1, it can be seen that the steady-state recognition method based on multi-factor multi-dimensional features of the cold atom absolute gravimeter provided by the embodiment of the present invention has a very high recognition accuracy rate.

[0047] Table 1 Test Results

[0048] A hull steady-state recognition method based on multi-factor multi-dimensional features of the cold atom absolute gravimeter (referred to as the multi-factor multi-dimensional feature method) provided by the embodiment of the present invention and a traditional recognition method based on single-factor single-dimensional features (referred to as the single-dimensional method), their recognition results for 4 kinds of hull steady states are as Figure 4 shown. Based on the verification results of the above embodiments, it can be known that the 7 kinds of hull steady-state feature elements used in the embodiment of the present invention, such as the X / Y / Z axis acceleration features, X / Y axis inclination features, and total atomic number and three-state atomic number features, the recognition accuracy rates using the above single features are 29.7%, 46.8%, 45.2%, 29.4%, 46.8%, 33.5%, 44.3% respectively, which are far lower than 98.9% of the multi-factor multi-feature method in the embodiment of the present invention. Therefore, the hull steady-state recognition method provided by the present invention has significant superiority compared with the traditional recognition method.

[0049] Figure 5 The structural diagram of a steady-state recognition system based on multi-factor multi-dimensional features of the cold atom absolute gravimeter provided by the embodiment of the present invention is as Figure 5As shown, a steady-state recognition system based on multi-element multi-dimensional features of a cold-atom absolute gravimeter includes an acquisition module, a processing module, a construction module, and an identification module, where: The acquisition module is used to collect historical data of the hull operation based on the cold-atom absolute gravimeter at each steady-state level, and the historical data includes the acceleration and inclination of the hull, as well as the atomic number data of the cold-atom absolute gravimeter; The processing module is used to divide the acceleration, inclination, and atomic number data into feature groups to form a multi-element multi-dimensional hull steady-state feature matrix; The construction module is used to establish a hull steady-state recognition model based on the hull steady-state feature matrix and machine learning algorithms, and perform cross-validation on the hull steady-state recognition model; The identification module is used to identify the hull steady state based on the acceleration, inclination, and atomic number data of the currently collected hull according to the hull steady-state recognition model after cross-validation.

[0050] It can be understood that a steady-state recognition system based on multi-element multi-dimensional features of a cold-atom absolute gravimeter provided by the present invention corresponds to the steady-state recognition method based on multi-element multi-dimensional features of a cold-atom absolute gravimeter provided in the foregoing embodiments. The relevant technical features of the steady-state recognition system based on multi-element multi-dimensional features of a cold-atom absolute gravimeter can refer to the relevant technical features of the steady-state recognition method based on multi-element multi-dimensional features of a cold-atom absolute gravimeter, which will not be elaborated here.

[0051] Please refer to Figure 6 , Figure 6 which is a schematic diagram of an embodiment of an electronic device provided by an embodiment of the present invention. As Figure 6 shown, an embodiment of the present invention provides an electronic device 600, including a memory 610, a processor 620, and a computer program 611 stored on the memory 610 and executable on the processor 620. When the processor 620 executes the computer program 611, the following steps are implemented: S1, at each steady-state level, collect historical data of the hull operation based on the cold-atom absolute gravimeter, and the historical data includes the acceleration and inclination of the hull, as well as the atomic number data of the cold-atom absolute gravimeter; S2, divide the acceleration, inclination, and atomic number data into feature groups to form a multi-element multi-dimensional hull steady-state feature matrix; S3, based on the hull steady-state feature matrix and machine learning algorithms, establish a hull steady-state recognition model, and perform cross-validation on the hull steady-state recognition model; S4, according to the hull steady-state recognition model after cross-validation, identify the hull steady state based on the acceleration, inclination, and atomic number data of the currently collected hull.

[0052] Please refer toFigure 7 , Figure 7 is a schematic diagram of an embodiment of a computer-readable storage medium provided by the present invention. As Figure 7 shown, this embodiment provides a computer-readable storage medium 700, on which a computer program 711 is stored. When the computer program 711 is executed by a processor, the following steps are implemented: S1. At each steady state level, based on the cold atom absolute gravimeter, collect historical data of the hull operation. The historical data includes the acceleration, inclination angle of the hull, and the atomic number data of the cold atom absolute gravimeter; S2. Divide the acceleration, inclination angle, and atomic number data into feature groups to form a multi-element and multi-dimensional hull steady state feature matrix; S3. Based on the hull steady state feature matrix and machine learning algorithms, establish a hull steady state recognition model, and perform cross-validation on the hull steady state recognition model; S4. According to the hull steady state recognition model after cross-validation, based on the currently collected acceleration, inclination angle, and atomic number data of the hull, identify the hull steady state.

[0053] A steady state recognition method, system, and storage medium based on multi-element and multi-dimensional features of a cold atom absolute gravimeter provided by an embodiment of the present invention have the following advantages: 1. The present invention can use the multi-element features of the cold atom absolute gravimeter to discriminate the hull steady state, which is more accurate than traditional single-element and single-dimensional features (such as a single inclination angle feature, acceleration feature, or attitude feature at a certain moment, etc.); 2. The present invention divides and discriminates multi-element features by groups, and identifies the hull steady state by instantaneously discriminating data of a specific quantity or duration in the past. This quantity or duration far exceeds the hull's swaying and settlement cycles, and the judgment result is more continuous and reliable; 3. The present invention first introduces the atomic number of the cold atom absolute gravity, which is vulnerable to the hull attitude, as a hull steady state discrimination feature, enhancing the diversity of discrimination features and the accuracy of the results.

[0054] It should be noted that in the above embodiments, the descriptions of each embodiment have their own emphases. For parts not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0055] Those skilled in the art should understand that the embodiments of the present invention can be provided as a method, a system, or a computer program product. Therefore, the present invention can adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present invention can adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk memories, CD-ROMs, optical memories, etc.) containing computer-usable program codes.

[0056] The present invention will be described with reference to the flowcharts and / or block diagrams of methods, apparatuses (systems), and computer program products according to embodiments of the present invention. It should be understood that each flow and / or block in the flowcharts and / or block diagrams, as well as the combination of flows and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to the processors of general-purpose computers, special-purpose computers, embedded computers, or other programmable data processing devices to generate a machine, such that the instructions executed by the processors of the computer or other programmable data processing devices generate means for implementing the functions specified in one or more flows Figure 1 one or more flows and / or blocks Figure 1 or one or more blocks.

[0057] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, such that the instructions stored in the computer-readable memory generate a manufactured article including instruction means that implement the functions specified in one or more flows Figure 1 one or more flows and / or blocks Figure 1 or one or more blocks.

[0058] These computer program instructions can also be loaded onto a computer or other programmable data processing device, such that a series of operational steps are executed on the computer or other programmable device to generate a computer-implemented process, and thus the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in one or more flows Figure 1 one or more flows and / or blocks Figure 1 or one or more blocks.

[0059] Although the preferred embodiments of the present invention have been described, those skilled in the art can make additional changes and modifications once they learn the basic creative concepts. Therefore, the appended claims are intended to be construed as including the preferred embodiments as well as all changes and modifications falling within the scope of the present invention.

[0060] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalent technologies, the present invention is also intended to include these modifications and variations.

Claims

1. A steady-state recognition method based on multi-element and multi-dimensional features of a cold-atom absolute gravimeter, characterized in that, Including the following steps: S1. At each steady state level, based on the cold atom absolute gravimeter, collect the historical data of the hull operation. The historical data includes the acceleration and inclination angle of the hull, as well as the atomic number data of the cold atom absolute gravimeter; S2. Divide the acceleration, inclination angle and atomic number data into feature groups to form a multi-element and multi-dimensional hull steady state feature matrix; S3. Based on the hull steady state feature matrix and the machine learning algorithm, establish a hull steady state recognition model and perform cross-validation on the hull steady state recognition model; S4. According to the hull steady state recognition model after cross-validation, based on the currently collected acceleration, inclination angle and atomic number data of the hull, identify the hull steady state.

2. The steady-state recognition method based on multi-factor multi-dimensional features of a cold atom absolute gravimeter according to claim 1, characterized in that, Step S1 includes: Based on the cold atom absolute gravimeter, collect the three-axis acceleration data of the X, Y, and Z axes of the hull operation at each steady state level; Based on the cold atom absolute gravimeter, collect the inclination angle states of the X and Y axes of the hull operation at each steady state level; According to the interference and state selection output results of cold atoms, collect the total atomic number and three-state atomic number of the cold atom absolute gravimeter of the hull operation at each steady state level.

3. A steady-state recognition method based on multi-factor multi-dimensional features of a cold atom absolute gravimeter according to claim 2, characterized in that, Step S2 includes: S201. For the data collected at the same steady state level, take the acceleration data of each axis, the inclination angle state data of each axis, the total atomic number and the three-state atomic number as a type of single-element feature data respectively. According to the preset dimension, divide each type of single-element feature data into multiple feature groups to form multiple single-element multi-dimensional steady state feature matrices; S202. Combine all the single-element multi-dimensional steady state feature matrices at the same steady state level to obtain a multi-element multi-dimensional steady state feature matrix at a single steady state level; S203. Combine the multi-element multi-dimensional steady state feature matrices at all steady state levels to form a hull steady state feature matrix.

4. A steady-state recognition method based on multi-factor multi-dimensional features of a cold-atom absolute gravimeter according to claim 3, characterized in that, In step S201, the step of dividing each type of single-element feature data into multiple feature groups according to the preset dimension to form multiple single-element multi-dimensional steady state feature matrices includes: Arrange a certain type of single-element feature data in columns; According to the preset number of rows or time period, divide the single-element feature data of the current type into multiple feature groups; Arrange the divided multiple feature groups in rows or columns to form a single-element multi-dimensional steady state feature matrix of the current type; Traverse all types of single-element feature data to obtain multiple single-element multi-dimensional steady state feature matrices.

5. A steady-state recognition method based on multi-element multi-dimensional features of a cold atom absolute gravimeter according to claim 3, characterized in that, In step S201, the step of dividing each type of single-element feature data into multiple feature groups according to the preset dimension to form multiple single-element multi-dimensional steady state feature matrices includes: Arrange a certain type of single-element feature data in rows; According to the preset number of columns or time period, divide the single-element feature data of the current type into multiple feature groups; Arrange the divided multiple feature groups in rows or columns to form a single-element multi-dimensional steady state feature matrix of the current type; Traverse all types of single-element feature data to obtain multiple single-element multi-dimensional steady state feature matrices.

6. A steady-state recognition method based on multi-element multi-dimensional features of a cold-atom absolute gravimeter according to any one of claims 3 to 5, characterized in that In step S1, it further includes: collecting the hydrological and meteorological feature data of the water area where the ship is located. The hydrological and meteorological feature data includes one or more of the course and speed, flow direction and velocity, wind direction and speed, and surge condition; In step S201, it further includes: For the hydrological and meteorological characteristic data collected at the same steady state level, according to the number of items of the collected hydrological and meteorological characteristic data, the hydrological and meteorological characteristic data are used as one or more types of single-element characteristic data; According to a preset dimension, each type of single-element characteristic data is divided into multiple characteristic groups to form a single or multiple single-element multi-dimensional steady state characteristic matrices corresponding to the number of items of the hydrological and meteorological characteristic data.

7. A steady-state recognition method based on multi-element and multi-dimensional features of a cold-atom absolute gravimeter according to claim 4 or 5, characterized in that In step S3, the machine learning algorithm is one of the linear discriminant analysis algorithm, support vector machine algorithm, random forest algorithm, neural network algorithm, and partial least squares algorithm; the cross-validation method adopted is K-fold cross-validation, leave-one-out cross-validation, or leave-K cross-validation.

8. A steady-state recognition system based on multi-factor multi-dimensional features of a cold-atom absolute gravimeter, characterized in that, Including: An acquisition module, configured to collect historical data of the hull operation based on a cold atom absolute gravimeter at each steady state level, where the historical data includes the acceleration and inclination angle of the hull, and the atomic number data of the cold atom absolute gravimeter; A processing module, configured to divide the acceleration, inclination angle, and atomic number data into characteristic groups to form a multi-element multi-dimensional hull steady state characteristic matrix; A construction module, configured to establish a hull steady state recognition model based on the hull steady state characteristic matrix and a machine learning algorithm, and perform cross-validation on the hull steady state recognition model; An identification module, configured to identify the hull steady state based on the acceleration, inclination angle, and atomic number data of the currently collected hull according to the hull steady state recognition model after cross-validation.

9. An electronic device, characterized in that, Including a memory and a processor, the processor is configured to implement the steps of the steady state recognition method based on the multi-element multi-dimensional characteristics of the cold atom absolute gravimeter according to any one of claims 1-7 when executing the computer management program stored in the memory.

10. A computer-readable storage medium, characterized in that, Stored thereon is a computer management program, and when the computer management program is executed by the processor, the steps of the steady state recognition method based on the multi-element multi-dimensional characteristics of the cold atom absolute gravimeter according to any one of claims 1-7 are implemented.

Images