Acceleration monitoring and forecasting method for sailing ship and ocean platform

Abstract

The invention provides an acceleration monitoring and forecasting method for a sailing ship and an ocean platform. The acceleration monitoring and forecasting method comprises the following steps: processing original sensor measurement data, carrying out operations such as filtering and differentiating on the data, and forecasting future motion data and motion amplitude. On the basis of general acceleration and angle sensor monitoring, a set of acceleration monitoring and calculating method is deduced, acceleration information of any point can be calculated according to acceleration and angle data measured by fixed points on a ship, and the method has the advantages of being simple in arrangement, clear in logic, complete in process and the like.

Description

Technical Field

[0001] The invention relates to an acceleration monitoring and forecasting method for sailing ships and marine platforms. Background Art

[0002] Acceleration monitoring and estimation methods are usually used to monitor the motion status of ships and offshore platforms, assess the degree of danger of ships and offshore platforms, and can usually be used in ship monitoring decision-making systems.

[0003] The development of ship-borne motion monitoring and decision-making systems is still in a blank stage in China. How to design a set of ship acceleration monitoring and calculation methods with complete functions, reliable data and easy deployment based on the current characteristics of domestic shipping vessels is an urgent problem to be solved. Summary of the invention

[0004] The object of the present invention is to provide an acceleration monitoring and prediction method for sailing ships and offshore platforms.

[0005] In order to solve the above problems, the present invention provides an acceleration monitoring and prediction method for sailing ships and offshore platforms, comprising:

[0006] Setting a posture sensor on a sailing ship or an offshore platform to obtain measurement data collected by the posture sensor on the sailing ship or the offshore platform;

[0007] Performing validity detection, filtering and differentiation processing on the measurement data to obtain processed data;

[0008] Based on the processed data, the acceleration at the metacenter of the sailing ship or offshore platform is calculated;

[0009] Based on the acceleration at the metacenter, the acceleration of any point to be measured on a sailing ship or an offshore platform can be calculated;

[0010] Training deep learning forecasting models based on the wave history and motion history of sailing ships or offshore platforms;

[0011] The trained deep learning prediction model can be used to predict the future motion data and motion amplitude of sailing ships or ocean platforms.

[0012] Furthermore, in the above method, a posture sensor is provided on a sailing ship or an ocean platform, including:

[0013] On a sailing ship or an offshore platform, four groups of attitude sensors are arranged, each group of attitude sensors includes: two three-axis acceleration sensors and one two-axis inclination sensor.

[0014] Furthermore, in the above method, a posture sensor is provided on a sailing ship or an ocean platform, including:

[0015] The posture sensors are arranged according to the following rules:

[0016] 1) Arranged at the far end of ships and offshore platforms;

[0017] 2) Arranged at a location far away from the metacenter;

[0018] 3) The linear acceleration sensors or inclination sensors are not arranged in a straight line;

[0019] 4) Arrange at least 4 linear acceleration sensors.

[0020] Furthermore, in the above method, the measurement data is subjected to validity detection, filtering and differentiation processing to obtain processed data.

[0021] The effectiveness detection includes:

[0022] Range detection includes: performing range detection on the input measurement data, that is, judging whether the measurement data falls within the current range according to the range interval set by the user;

[0023] Variance detection includes: detecting the variance of the input measurement data, that is, judging whether the variance of the measurement data falls within the current range according to the range set by the user;

[0024] Bad pixel detection includes: judging whether there are data points in the measured data that are beyond the normal range according to the range interval set by the user.

[0025] Furthermore, in the above method, the bad pixel detection adopts 3-σ criterion.

[0026] Furthermore, in the above method, the measurement data is subjected to validity detection, filtering and differentiation processing to obtain processed data, including:

[0027] Performing validity detection on the measurement data;

[0028] The measured data that has passed the validity detection is filtered and differentiated, where:

[0029] The filtering adopts a real-time filtering algorithm, which depends on the filtering results and measurement values ​​of the previous two steps at the current moment;

[0030] Differentiation uses two-point, three-point or five-point interpolation differentiation formulas in numerical analysis.

[0031] Furthermore, in the above method, the filtering adopts a Butterworth filter.

[0032] Furthermore, in the above method, based on the processed data, the acceleration calculation at the metacenter of the sailing ship or the offshore platform is performed, including:

[0033] Based on the processed data, the angular velocity, angular acceleration and linear acceleration of the entire ship are calculated;

[0034] Based on the angular velocity, angular acceleration and linear acceleration of the entire ship, the acceleration at the metacenter of the sailing ship or offshore platform is obtained.

[0035] Furthermore, in the above method, based on the angular velocity, angular acceleration and linear acceleration of the whole ship, the acceleration at the metacenter of the sailing ship or the offshore platform is obtained, including:

[0036] Based on the processed data, the angular velocity, angular acceleration and linear acceleration of the entire ship are calculated, including:

[0037] Based on the processed data corresponding to each linear acceleration sensor, the whole ship angular velocity, angular acceleration and linear acceleration of each point corresponding to each linear acceleration sensor are calculated;

[0038] Based on the angular velocity, angular acceleration and linear acceleration of the whole ship, the acceleration at the metacenter of the sailing ship or offshore platform is obtained, including:

[0039] The whole ship angular velocity, angular acceleration and linear acceleration of each point corresponding to each linear acceleration sensor are used to obtain the acceleration at the metacenter of the sailing ship or offshore platform corresponding to each linear acceleration sensor;

[0040] The acceleration at the metacenter of the sailing ship or the offshore platform corresponding to each linear acceleration sensor is data-fused to obtain the final acceleration at the metacenter of the sailing ship or the offshore platform.

[0041] Furthermore, in the above method, the deep learning prediction model adopts an LSTM model.

[0042] Compared with the prior art, the present invention has the following advantages:

[0043] 1. According to different acceleration and angle sensor layout schemes, there can be corresponding acceleration monitoring and calculation algorithms, which have high flexibility in sensor layout and installation.

[0044] 2. By adopting a redundant sensor layout, the same data can have multiple backups and can be compared with each other, which improves the reliability and accuracy of the data.

[0045] 3. The core algorithm of acceleration monitoring and calculation is complete. In addition to acceleration calculation, there are also various data processing functions, including data validity monitoring function, filtering function, differential function, etc. The data processing is relatively complete and the logic is clear.

[0046] 4. Combined with the currently widely used artificial neural network, it can further predict the future status of the ship, and retains space for adding other functions, with strong scalability. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 is a side view of a posture sensor arrangement according to an embodiment of the present invention;

[0048] Figure 2 is a top view of the arrangement of a posture sensor according to an embodiment of the present invention;

[0049] Figure 3 It is a schematic diagram of data flow of an acceleration monitoring and prediction method for sailing ships and marine platforms according to an embodiment of the present invention;

[0050] Figure 4 is a schematic diagram of a neural network prediction model according to an embodiment of the present invention;

[0051] Figure 5 is a schematic diagram of data flow related functions according to an embodiment of the present invention;

[0052] Figure 6 1 is a schematic diagram of an RNN network model according to an embodiment of the present invention;

[0053] Figure 7 It is a schematic diagram of the RNN, LSTM, and GRU structures of an embodiment of the present invention;

[0054] Figure 8 It is a schematic diagram of the structure of a ship state prediction model according to an embodiment of the present invention;

[0055] Fig. 9 is a schematic diagram of range detection according to an embodiment of the present invention;

[0056] Fig.10 is a schematic diagram of variance detection according to an embodiment of the present invention;

[0057] Fig.11 Schematic diagram of bad pixel detection according to an embodiment of the present invention

[0058] Fig.12 2 is a filtering schematic diagram of an embodiment of the present invention. DETAILED DESCRIPTION

[0059] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0060] like Figure 1 As shown, the present invention provides an acceleration monitoring and prediction method for sailing ships and marine platforms, comprising:

[0061] Step S1, setting a posture sensor on a sailing ship or an offshore platform, and acquiring measurement data collected by the posture sensor on the sailing ship and the offshore platform;

[0062] Step S2, performing validity detection, filtering and differentiation processing on the measurement data to obtain processed data;

[0063] Step S3, calculating the acceleration at the metacenter of the sailing ship or the offshore platform based on the processed data;

[0064] Step S4, based on the acceleration at the metacenter, calculating the acceleration of any point to be measured on the sailing ship or the offshore platform;

[0065] Step S5, training a deep learning prediction model based on the wave history and motion history of the sailing ship or the offshore platform;

[0066] Step S6, predicting the future motion data and motion amplitude of the sailing ship or ocean platform through the trained deep learning prediction model.

[0067] Here, the present invention mainly consists of the following parts: 1) acceleration monitoring equipment of attitude sensor; 2) acceleration, angle data processing and acceleration estimation system; 3) neural network motion prediction system. The present invention can be used to monitor the state conditions such as ship motion acceleration, roll angle and pitch angle, and can predict the ship motion in the future according to the current ship state and marine environment, so as to provide reference and help for crew members to understand the ship state and make decisions.

[0068] The present invention derives a set of acceleration monitoring and calculation methods based on universal acceleration and angle sensor monitoring, which can realize the calculation of acceleration information of any point from the acceleration and angle data measured at fixed points on the ship, and has the advantages of simple layout, clear logic, complete process, etc.

[0069] In one embodiment of the acceleration monitoring and prediction method for sailing ships and marine platforms of the present invention, step S1, setting a posture sensor on the sailing ship or the marine platform, comprises:

[0070] On a sailing ship or an offshore platform, four groups of attitude sensors are arranged, each group of attitude sensors includes: two three-axis acceleration sensors and one two-axis inclination sensor.

[0071] Here, the acceleration angle sensor arrangement may involve a total of 4 groups of attitude sensors, each group of attitude sensors includes 2 triaxial acceleration sensors and 1 biaxial inclination sensor, which can be arranged in the front and rear of the pipe gallery in the bottom of the ship, and in the empty compartments on both sides, such as Figure 1 and Figure 2 shown.

[0072] By arranging the attitude sensors as above, the three-dimensional linear acceleration and angle change data at four positions on the ship can be obtained, and then these data can be processed to obtain reliable measurement data.

[0073] In one embodiment of the acceleration monitoring and prediction method for sailing ships and marine platforms of the present invention, step S1, setting a posture sensor on the sailing ship or the marine platform, comprises:

[0074] The posture sensors are arranged according to the following rules:

[0075] 1) Arranged at the far end of ships and offshore platforms;

[0076] 2) Arranged at a location far away from the metacenter;

[0077] 3) The linear acceleration sensors or inclination sensors are not arranged in a straight line;

[0078] 4) Arrange at least 4 linear acceleration sensors.

[0079] Specifically, assuming that only four linear acceleration sensors are arranged, the acceleration at the metacenter can be derived in the following way.

[0080] Assume that the instantaneous rotation center of the ship is (x 0 ,y 0 , z 0 ), the positions of the four linear accelerations are (x 1 ,y 1 ,z 1 ), (x 2 ,y 2 ,z 2 ),(x 3 ,y 3 ,z 3 ),(x 4 ,y 4 ,z 4 ). Then, the radius vectors of the four linear accelerations relative to the instantaneous rotation center are

[0081]

[0082] Assume that the roll, pitch and bow angles of the ship are (φ, θ, ψ), and the accelerations of surge, sway and heave are Assume that the acceleration values ​​of the four linear accelerations are

[0083]

[0084] Then, according to the acceleration synthesis theory in theoretical mechanics, we can get

[0085]

[0086] Where, the angular velocity and angular acceleration are

[0087]

[0088] In the formula, when expanded in three directions (x, y, z), there are 12 equations and 12 unknowns. Therefore the problem is completely solvable.

[0089] Assuming that in addition to the linear acceleration sensor, an additional angle sensor is also arranged, then the above formula The data measured by the angle sensor can be directly obtained through data processing operations such as differential filtering.

[0090] Then, the acceleration at the metacenter of each single-line acceleration sensor can be derived in the following way.

[0091] If we consider the rotation in one direction alone, taking pitch as an example, we can Decompose into So

[0092]

[0093]

[0094] but

[0095]

[0096] Can write

[0097]

[0098] but

[0099]

[0100] The above are the relevant principles and theories of sensor arrangement. By arranging the sensors in this way, the acceleration at the metacenter can be derived.

[0101] In one embodiment of the acceleration monitoring and prediction method for sailing ships and marine platforms of the present invention, step S2 performs validity detection, filtering and differential processing on the measurement data to obtain processed data.

[0102] The effectiveness detection includes:

[0103] Range detection includes: performing range detection on the input measurement data, that is, judging whether the measurement data falls within the current range according to the range interval set by the user;

[0104] Variance detection includes: detecting the variance of the input measurement data, that is, judging whether the variance of the measurement data falls within the current range according to the range set by the user;

[0105] Bad pixel detection includes: judging whether there are data points beyond the normal range according to the range interval set by the user, usually using 3-σ accuracy.

[0106] Specifically, Figure 3 The various processes in the main description of the input and output of data, Figure 5 The data flow and processing of each intermediate part are introduced in more detail.

[0107] The validity check function receives two input variables, data and PARAMS, and outputs the data of the sensors that passed the test. For example, if 4 sets of linear acceleration and angle data of sensors are input, and one set fails the validity check, the data of the remaining 3 sets of sensors will be returned.

[0108] Range detection can be done as Fig. 9 As shown; the variance detection module can be Fig.10 As shown; bad pixel detection can be as follows Fig.11 shown.

[0109] In one embodiment of the acceleration monitoring and prediction method for sailing ships and marine platforms of the present invention, step S2, performing validity detection, filtering and differential processing on the measurement data to obtain processed data, includes:

[0110] Performing validity detection on the measurement data;

[0111] The measured data that has passed the validity detection is filtered and differentiated, where:

[0112] like Fig.12 As shown, the filtering adopts a real-time filtering algorithm, which depends on the filtering results and measurement values ​​of the previous two steps at the current moment;

[0113] Differentiation uses two-point, three-point or five-point interpolation differentiation formulas in numerical analysis.

[0114] Among them, the three-point interpolation differential formula is:

[0115]

[0116] The five-point interpolation differential formula is:

[0117]

[0118] Optionally, in the present invention, four groups of sensors may be provided, each group including two three-axis acceleration sensors and one two-axis inclination sensor, so that the sensor data will be tested for validity upon input, and sensor data that fails the test will be discarded. Usually, failure of the validity test may be due to problems with the sensor itself, or errors in sensor installation, etc. Redundant acceleration sensors and angle sensors can still provide acceleration and angle data at corresponding positions when sensors in the same group fail, and can fuse similar data to make the data more accurate and reliable. Redundant sensor arrangement is to provide corresponding functions when one or more sensors fail.

[0119] The Butterworth filter (BF) can be used, which is a low-pass filter suitable for suppressing high-frequency noise. In addition, in actual implementation, a real-time filtering method is used, that is, the filtering of each data point depends on its first two filtered values ​​and the observed value. This method has been proven to be effective in operation.

[0120] In one embodiment of the acceleration monitoring and prediction method for sailing ships and offshore platforms of the present invention, step S3, based on the processed data, calculates the acceleration at the metacenter of the sailing ship or offshore platform, including:

[0121] Based on the processed data, the angular velocity, angular acceleration and linear acceleration of the entire ship are calculated;

[0122] Based on the angular velocity, angular acceleration and linear acceleration of the entire ship, the acceleration at the metacenter of the sailing ship or offshore platform is obtained.

[0123] Specifically, after passing the validity test, these data are filtered and differentiated, and the processed data of the same type will be fused, such as the acceleration data at the same point and the angle data of the entire ship.

[0124] After preliminary calculation of these data, some output data can be obtained, such as the angular velocity of the whole ship, angular acceleration, acceleration of each point line, etc. Further calculation of the above data, mainly through the acceleration calculation formula derived in the previous article, can obtain the acceleration value at the metacenter.

[0125] In one embodiment of the acceleration monitoring and prediction method for sailing ships and offshore platforms of the present invention, based on the angular velocity, angular acceleration and the acceleration of each point line of the whole ship, the acceleration at the metacenter of the sailing ship or the offshore platform is obtained, including:

[0126] Based on the processed data, the angular velocity, angular acceleration and linear acceleration of the entire ship are calculated, including:

[0127] Based on the processed data corresponding to each linear acceleration sensor, the whole ship angular velocity, angular acceleration and linear acceleration of each point corresponding to each linear acceleration sensor are calculated;

[0128] Based on the angular velocity, angular acceleration and linear acceleration of the whole ship, the acceleration at the metacenter of the sailing ship or offshore platform is obtained, including:

[0129] The whole ship angular velocity, angular acceleration and linear acceleration of each point corresponding to each linear acceleration sensor are used to obtain the acceleration at the metacenter of the sailing ship or offshore platform corresponding to each linear acceleration sensor;

[0130] The acceleration at the metacenter of the sailing ship or the offshore platform corresponding to each linear acceleration sensor is data-fused to obtain the final acceleration at the metacenter of the sailing ship or the offshore platform.

[0131] Specifically, since there are multiple linear acceleration sensors, the acceleration value at the metacenter corresponding to each sensor can be calculated based on the processed data of each linear acceleration sensor. After data fusion of these values, the final acceleration value at the metacenter can be obtained as the acceleration data of the entire ship. And through the acceleration data of the entire ship and the coordinates of the dangerous monitoring points of concern, the acceleration of any point to be measured on the entire ship can be obtained. The final functions and data processing process are as follows: Figure 6 The consistency involved.

[0132] In one embodiment of the acceleration monitoring and prediction method for sailing ships and marine platforms of the present invention, the deep learning prediction model adopts an LSTM model.

[0133] Specifically, by establishing a deep learning prediction model, the ship's motion within several periods can be obtained based on the ship's motion history data and ocean environmental conditions.

[0134] According to the characteristics of ship and ocean platform motion prediction problems, a recurrent neural network model suitable for processing time series data can be selected as the basis of the prediction model. Recurrent Neural Network (RNN) is a special type of neural network algorithm that has great advantages in processing time series data. Usually, the input and output of a neural network model correspond one to one, and there is no connection between the two adjacent inputs and outputs. However, for some data with actual sequence relationships, this information will be ignored and lost during processing.

[0135] In RNN, by associating the weights of the intermediate layers with the previous output, sequence information can be better processed. Figure 6In the example, X is the input, S is the value of the middle layer, O is the output, u is the weight from the input layer to the middle layer, v is the weight from the middle layer to the output layer, and w is the weight from the previous middle layer to the next middle layer. The existence of w makes each adjacent input connected, that is, the value of the middle layer at time t depends not only on the input at time t, but also on the value of the middle layer at time t-1.

[0136] The calculation formula of the classic RNN network model is:

[0137] S t =f(W·S t-1 +U·X t )

[0138] O t =g(V·S t )

[0139] The classic RNN structure has the problems of gradient vanishing and gradient exploding, and it is difficult to get good results in actual training. Therefore, at the end of the last century, some experts and scholars adjusted and involved the RNN model structure and proposed many RNN variants, such as the most widely used long short-term memory network (LSTM) in RNN, and the gated recurrent neural network (GRU), whose structure is as follows Figure 7 As shown. The differences and connections between LSTM and GRU are mainly:

[0140] 1) The performance of LSTM and GRU is comparable in many tasks;

[0141] 2) GRU has fewer parameters, 1 / 3 fewer parameters, so it is easier to converge, but in the case of large data sets, LSTM performs better;

[0142] 3) GRU has only two gates (update, reset), LSTM has three gates (forget, input, output). GRU directly passes the hidden state to the next unit, while LSTM uses memory cell to package the hidden state.

[0143] Both can be used as choices for model building, and existing mature third-party libraries contain their original models. Considering that modeling accuracy and prediction accuracy are the first criteria for measurement, it is appropriate to use the LSTM model for prediction.

[0144] In predicting the movement of ships and offshore platforms, the expected input is the wave history and the movement history, both of which are time series data, and in actual situations, the movement response at a certain moment must be the result of the movement and waves at the previous moment, and has a certain continuity. Therefore, it is reasonable and accurate to use RNN as an algorithm to establish a forecast model, and it has a good performance in processing relevant time series data.

[0145] like Figure 4 As described in , the input of the forecast model uses environmental condition data (such as wave history data) and motion history data (such as vertical and vertical motion), and the dimension of the input layer is 2. The type of RNN layer can be LSTM or GRU, and the fully connected layer is used as the final output. The data flow is as follows Figure 8 shown.

[0146] The subsequent functional expansion is mainly based on the current state of the ship, such as the current ship motion, acceleration, angle change, and the monitored ocean wave environmental conditions, etc., to calculate and predict the movement of the ship in the future through a certain method, and can predict the maximum movement amplitude of the ship in the future, which can be used as a reference for the crew, and based on this, timely adjust the ship's posture.

[0147] By adopting the above scheme, the acceleration monitoring and estimation method proposed in the present invention has the following characteristics:

[0148] 1. According to different acceleration and angle sensor layout schemes, there can be corresponding acceleration monitoring and calculation algorithms, which have high flexibility in sensor layout and installation.

[0149] 2. By adopting a redundant sensor layout, the same data can have multiple backups and can be compared with each other, which improves the reliability and accuracy of the data.

[0150] 3. The core algorithm of acceleration monitoring and calculation is complete. In addition to acceleration calculation, there are also various data processing functions, including data validity monitoring function, filtering function, differential function, etc. The data processing is relatively complete and the logic is clear.

[0151] 4. Combined with the currently widely used artificial neural network, it can further predict the future status of the ship, and retains space for adding other functions, with strong scalability.

[0152] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referenced to each other.

[0153] Professionals may further appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in the above description according to function. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians may use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0154] Obviously, those skilled in the art can make various changes and modifications to the invention without departing from the spirit and scope of the invention. Thus, if these modifications and variations of the invention fall within the scope of the claims of the invention and their equivalents, the invention is also intended to include these modifications and variations.

Claims

1. A method for monitoring and predicting acceleration of sailing ships and offshore platforms, It is characterized in that include: Setting a posture sensor on a sailing ship or an offshore platform to obtain measurement data collected by the posture sensor on the sailing ship or the offshore platform; Performing validity detection, filtering and differentiation processing on the measurement data to obtain processed data; Based on the processed data, the acceleration at the metacenter of the sailing ship or offshore platform is calculated; Based on the acceleration at the metacenter, the acceleration of any point to be measured on a sailing ship or an offshore platform can be calculated; Training deep learning forecasting models based on the wave history and motion history of sailing ships or offshore platforms; The trained deep learning prediction model can be used to predict the future motion data and motion amplitude of sailing ships or ocean platforms.

2. The acceleration monitoring and prediction method for sailing ships and offshore platforms according to claim 1, It is characterized in that The attitude sensor is set on a sailing ship or an offshore platform, including: On a sailing ship or an offshore platform, four groups of attitude sensors are arranged, each group of attitude sensors includes: two three-axis acceleration sensors and one two-axis inclination sensor.

3. The acceleration monitoring and prediction method for sailing ships and offshore platforms according to claim 1, It is characterized in that The attitude sensor is set on a sailing ship or an offshore platform, including: The posture sensors are arranged according to the following rules: 1) Arranged at the far end of ships and offshore platforms; 2) Arranged at a location far away from the metacenter; 3) The linear acceleration sensors or inclination sensors are not arranged in a straight line; 4) Arrange at least 4 linear acceleration sensors.

4. The acceleration monitoring and prediction method for sailing ships and offshore platforms according to claim 1, It is characterized in that The measurement data is subjected to validity detection, filtering and differentiation processing to obtain processed data. The effectiveness detection includes: Range detection includes: performing range detection on the input measurement data, that is, judging whether the measurement data falls within the current range according to the range interval set by the user; Variance detection includes: detecting the variance of the input measurement data, that is, judging whether the variance of the measurement data falls within the current range according to the range set by the user; Bad pixel detection includes: judging whether there are data points in the measured data that are beyond the normal range according to the range interval set by the user.

5. The acceleration monitoring and prediction method for sailing ships and offshore platforms as claimed in claim 4, It is characterized in that The bad pixel detection adopts 3-σ standard measurement and judgment.

6. The acceleration monitoring and prediction method for sailing ships and offshore platforms according to claim 1, It is characterized in that The measurement data is subjected to validity detection, filtering and differentiation processing to obtain processed data, including: Performing validity detection on the measurement data; The measured data that has passed the validity detection is filtered and differentiated, where: The filtering adopts a real-time filtering algorithm, which depends on the filtering results and measurement values ​​of the previous two steps at the current moment; Differentiation uses two-point, three-point or five-point interpolation differentiation formulas in numerical analysis.

7. The acceleration monitoring and prediction method for sailing ships and offshore platforms according to claim 6, It is characterized in that The filtering method is a Butterworth filter.

8. The acceleration monitoring and prediction method for sailing ships and offshore platforms according to claim 1, It is characterized in that Based on the processed data, the acceleration at the metacenter of the sailing ship or offshore platform is calculated, including: Based on the processed data, the angular velocity, angular acceleration and linear acceleration of the entire ship are calculated; Based on the angular velocity, angular acceleration and linear acceleration of the entire ship, the acceleration at the metacenter of the sailing ship or offshore platform is obtained.

9. The acceleration monitoring and prediction method for sailing ships and offshore platforms according to claim 8, It is characterized in that Based on the angular velocity, angular acceleration and linear acceleration of the whole ship, the acceleration at the metacenter of the sailing ship or offshore platform is obtained, including: Based on the processed data, the angular velocity, angular acceleration and linear acceleration of the entire ship are calculated, including: Based on the processed data corresponding to each linear acceleration sensor, the angular velocity, angular acceleration and linear acceleration of each point of the whole ship corresponding to each linear acceleration sensor are calculated; Based on the angular velocity, angular acceleration and linear acceleration of the whole ship, the acceleration at the metacenter of the sailing ship or offshore platform is obtained, including: The whole ship angular velocity, angular acceleration and linear acceleration of each point corresponding to each linear acceleration sensor are used to obtain the acceleration at the metacenter of the sailing ship or offshore platform corresponding to each linear acceleration sensor; The acceleration at the metacenter of the sailing ship or the offshore platform corresponding to each linear acceleration sensor is data-fused to obtain the final acceleration at the metacenter of the sailing ship or the offshore platform.

10. The acceleration monitoring and prediction method for sailing ships and offshore platforms according to claim 1, It is characterized in that The deep learning prediction model adopts the LSTM model.

Images