Ship rolling prediction method

Abstract

The invention relates to the technical field of ship motion postures, in particular to a ship rolling prediction method, which comprises the following steps of S1, constructing predicted ship rolling data based on a ship rolling simulation model; s2, establishing a long-short-term memory network model, and determining parameters, needing to be optimized, of the model; s3, initializing particle swarm parameters; s4, determining a fitness function of the particle population through iterative training; s5, optimizing particle parameters; and S6, judging whether the maximum number of iterations is reached or not, if the maximum number of iterations is reached, transmitting the obtained optimal parameters to the long-short-term memory network model, and carrying out training and prediction, and if the requirements are not met, returning to the step S5. The rolling prediction model based on the particle swarm algorithm and the long-short-term memory network solves the training problem of initialization parameter configuration in an existing ship rolling prediction method.

Description

Technical Field

[0001] The present invention relates to the technical field of ship motion attitude, and in particular to a ship roll prediction method. Background Art

[0002] At present, as a means of transportation that mainly operates in water, ships are subject to the influence of factors such as marine weather and sea waves when performing various operations at sea, which causes the ships to produce six degrees of freedom movements in different directions, including sway, surge, roll, pitch, yaw and heave. Such degrees of freedom movements will seriously interfere with the ship's marine operations, especially the impact of roll on the ship's marine movement is particularly serious. Roll refers to the degree to which the ship shakes left and right with the longitudinal axis as the center.

[0003] Regarding the above-mentioned related technologies, it is found that in the research methods of ship roll prediction, traditional prediction methods include statistical prediction methods, convolution methods, Kalman filtering methods and periodogram methods. However, the above-mentioned roll prediction algorithms all require artificially specifying the values ​​or ranges of the initial parameters, which will make the prediction results of the final model unpredictable and will also cause the trained prediction model to be not optimal. Summary of the invention

[0004] The present invention solves the problems in the related art, proposes a ship roll prediction method, and solves the problem that the initialization parameter configuration of the existing ship roll prediction is inaccurate and affects the prediction result.

[0005] In order to solve the above technical problems, the present invention is implemented by the following technical solutions: A ship rolling prediction method comprises the following steps:

[0006] S1: Based on the ship rolling simulation model, the predicted ship rolling data is constructed;

[0007] S2: Establish a long short-term memory network model and determine the parameters that need to be optimized;

[0008] S3: Initialize particle population parameters;

[0009] S4: Determine the fitness function of the particle population through iterative training;

[0010] S5: particle parameter optimization;

[0011] S6: Determine whether the maximum number of iterations has been reached. If so, pass the optimal parameters to the long short-term memory network model for training and prediction. If not, return to step S5.

[0012] As a preferred solution, S1 includes: based on the ship roll simulation model, after Laplace transformation and normalization, setting the roll damping coefficient, generating a roll mathematical motion model, and generating roll simulation data by constructing a roll simulation curve, which is used as the data source for the model evaluation standard.

[0013] As a preferred solution, the parameters that need to be optimized for the model in S2 include: model learning rate, number of training times and number of neurons, wherein the number of neurons includes the number of hidden units in the first layer and the number of hidden units in the second layer.

[0014] As a preferred solution, the particle population parameters in S3 include: initial velocity, relative position, number of training times and initial weight of particles.

[0015] As a preferred solution, in S5, the MAPE value is used as the particle fitness function in the prediction model, and the optimal parameters of the model are found based on the MAPE value.

[0016] Compared with the prior art, the beneficial effects of the present invention are as follows: the present invention constructs predicted ship roll data based on a ship roll simulation model, combines a long short-term memory network model to determine the parameters affecting the model, initializes and determines the particle population related parameters, generates a fitness function through iterative training, and finally optimizes the parameters using the mean absolute percentage error as a model evaluation criterion, and then transmits the obtained optimal parameters to the long short-term memory network model for training and prediction. The present application solves the training problem of initialization parameter configuration in the existing ship roll prediction method based on a roll prediction model of a particle population algorithm and a long short-term memory network. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 is a flow chart of the model training algorithm of the present invention;

[0018] Figure 2 It is the training result of the long short-term memory network model in the present invention;

[0019] Figure 3 It is an iterative process of the long short-term memory network model based on the particle population algorithm of the present invention. DETAILED DESCRIPTION

[0020] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments. The following description of at least one exemplary embodiment is actually only illustrative and is by no means intended to limit the present invention and its application or use. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0021] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit the exemplary embodiments according to the present application. As used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, it should be understood that when the terms "comprise" and / or "include" are used in this specification, it indicates the presence of features, steps, operations, devices, components and / or combinations thereof.

[0022] Unless otherwise specifically stated, the relative arrangement of the parts and steps described in these embodiments, numerical expressions and numerical values ​​do not limit the scope of the present invention. At the same time, it should be understood that, for ease of description, the sizes of the various parts shown in the accompanying drawings are not drawn according to the actual proportional relationship. The technology, method and equipment known to ordinary technicians in the relevant field may not be discussed in detail, but in appropriate cases, the technology, method and equipment should be regarded as a part of the authorization specification. In all examples shown and discussed here, any specific value should be interpreted as being merely exemplary, rather than as a limitation. Therefore, other examples of exemplary embodiments may have different values. It should be noted that similar reference numerals and letters represent similar items in the following drawings, so once a certain item is defined in an accompanying drawing, it does not need to be further discussed in subsequent drawings.

[0023] In the description of the present invention, it is necessary to understand that the directions or positional relationships indicated by directional words such as "front, back, up, down, left, right", "lateral, vertical, perpendicular, horizontal" and "top, bottom" are usually based on the directions or positional relationships shown in the drawings. They are only for the convenience of describing the present invention and simplifying the description. Unless otherwise specified, these directional words do not indicate or imply that the devices or elements referred to must have a specific direction or be constructed and operated in a specific direction. Therefore, they cannot be understood as limiting the scope of protection of the present invention. The directional words "inside and outside" refer to the inside and outside relative to the contours of each component itself.

[0024] For ease of description, spatially relative terms such as "above", "above", "on the upper surface of", "above", etc. may be used here to describe the spatial positional relationship between a device or feature and other devices or features as shown in the figure. It should be understood that spatially relative terms are intended to include different orientations of the device in use or operation in addition to the orientation described in the figure. For example, if the device in the accompanying drawings is inverted, the device described as "above other devices or structures" or "above other devices or structures" will be positioned as "below other devices or structures" or "below other devices or structures". Thus, the exemplary term "above" may include both "above" and "below". The device may also be positioned in other different ways (rotated 90 degrees or in other orientations), and the spatially relative descriptions used here are interpreted accordingly.

[0025] In addition, it should be noted that the use of terms such as "first" and "second" to limit components is only for the convenience of distinguishing the corresponding components. If not otherwise stated, the above terms have no special meaning and therefore cannot be understood as limiting the scope of protection of the present invention.

[0026] Reference Figure 1 , Figure 2 and Figure 3 As shown, a ship roll prediction method, first, based on the ship roll simulation model, after Laplace transformation and normalization, the roll damping coefficient is set to generate a roll mathematical motion model, and the roll simulation data is generated by constructing a roll simulation curve, which is used as the data source of the model evaluation standard to construct the predicted ship roll data.

[0027] Next, a long short-term memory network model (LSTM) is established to determine the parameters that need to be optimized for the model, including the model learning rate, the number of training times, and the number of neurons, where the number of neurons includes the number of hidden units in the first layer and the number of hidden units in the second layer.

[0028] LSTM is a long short-term memory artificial neural network, a deep learning model commonly used to process sequence data. It is a special recurrent neural network.

[0029] Next, initialize the particle population parameters, including the particle's initial velocity, relative position, number of trainings, and initial weight. At runtime, a random initialization particle is first generated, and the parameters include the particle's initial velocity, relative position, number of trainings, and initial weight. Through iterative training, determine the fitness function of the particle population. In the process of continuous parameter adjustment of particles, dynamically find the optimal position and historical global optimal position of each particle, and update the optimal fitness value.

[0030] Next, particle parameter optimization is performed, and the MAPE value is used as the particle fitness function in the prediction model to find the optimal model parameters.

[0031] MAPE (Mean Absolute Percentage Error): Mean absolute percentage error is a measure commonly used to evaluate the error between the predicted value and the actual value. It is an important indicator to measure the accuracy of the predicted value. MAPE appears in the form of a percentage, indicating the average percentage of the relative error between the predicted value and the actual value. The value range of MAPE is [0, +∞), and the smaller the value, the more accurate the prediction model. Generally speaking, a MAPE less than 10% is considered to be a relatively good prediction model, and a MAPE between 10% and 20% indicates that the prediction accuracy is acceptable. However, if MAPE is greater than 20%, the prediction effect is not ideal, and the accuracy of the prediction model needs to be further improved. The specific calculation is the square root of the mean of the square of the error between the true value and the predicted value, and this indicator is used to evaluate the model construction.

[0032] Finally, it is determined whether the maximum number of iterations has been reached. If so, the optimal parameters are passed to the long short-term memory network model for training and prediction. If not, the particle parameter optimization is returned.

[0033] The number of iterations refers to the total number of times the model weights are updated during the training process, which is usually equal to the number of epochs multiplied by the number of batches per epoch. An epoch means that the entire training dataset is used for training once.

[0034] The repeated particle parameter optimization is to satisfy the PSO-LSTM combined model timing transmission algorithm so as to modify the weight each time according to the error, so that the data error can be repeatedly calculated, thereby minimizing the training error.

[0035] The PSO algorithm is an optimization algorithm based on swarm intelligence. It simulates the group behavior of birds or fish flocks, and finds the optimal solution to the problem through continuous iteration. In the PSO algorithm, each individual is regarded as a potential solution to the problem, and they can find a better solution by exchanging information with each other. Each individual has its own position and corresponding matching speed, and continuously adjusts its position by combining individual experience and group experience. The PSO algorithm has been successfully applied to solve a variety of optimization problems, such as neural network training, function optimization, combinatorial optimization, image processing, etc. In the specific iterative calculation process, the PSO algorithm continuously updates the position and speed of each individual and finds the optimal solution in the search space. The update of the position and speed of each individual is affected by the global optimal position and the optimal position of the individual in previous generations.

[0036] The above are preferred implementation modes of the present invention. Those skilled in the art to which the present invention belongs can also change and modify the above implementation modes. Therefore, the present invention is not limited to the above specific implementation modes. Any obvious improvements, substitutions or modifications made by those skilled in the art on the basis of the present invention belong to the protection scope of the present invention.

Claims

1. A ship rolling prediction method, characterized in that: The steps include: S1: Based on the ship rolling simulation model, the predicted ship rolling data is constructed; S2: Establish a long short-term memory network model and determine the parameters that need to be optimized; S3: Initialize particle population parameters; S4: Determine the fitness function of the particle population through iterative training; S5: particle parameter optimization; S6: Determine whether the maximum number of iterations has been reached. If so, pass the optimal parameters to the long short-term memory network model for training and prediction. If not, return to step S5.

2. A ship rolling prediction method according to claim 1, characterized in that: The S1 includes: based on the ship roll simulation model, after Laplace transformation and normalization, setting the roll damping coefficient, generating a roll mathematical motion model, and generating roll simulation data by constructing a roll simulation curve, which is used as a data source for the model evaluation standard.

3. A ship rolling prediction method according to claim 2, characterized in that: The parameters that need to be optimized for the model in S2 include: model learning rate, number of training times and number of neurons, where the number of neurons includes the number of hidden units in the first layer and the number of hidden units in the second layer.

4. A ship rolling prediction method according to claim 3, characterized in that: The particle population parameters in S3 include: initial velocity, relative position, number of training times and initial weight of particles.

5. A ship rolling prediction method according to claim 4, characterized in that: In S5, the MAPE value is used as the particle fitness function in the prediction model, and the optimal parameters of the model are found based on the MAPE value.

Images