Ship speed prediction and optimization method and device and readable storage medium

Abstract

The invention discloses a ship navigational speed prediction and optimization method and device and a readable storage medium, and the method comprises the steps: building a ship navigational speed prediction model by taking the host power, the ship course, the flow velocity, the relative flow direction, the wind speed, the relative wind direction, the effective wave height, the relative wave direction and the combined period of stormy waves and surges as the input, and taking the ship navigational speed as the output; acquiring ship route planning information and marine weather forecast information; dividing the planned ship route into a plurality of planned route segments; acquiring related information of each planned leg in combination with marine weather forecast information, and predicting real ship navigational speeds of each planned leg of the ship under corresponding weather sea conditions and different host powers in combination with a ship navigational speed prediction model; and establishing a cost function according to the host power and the corresponding real ship navigational speed, constructing a ship navigational speed optimization model in combination with the constraint condition, and finally obtaining the optimized navigational speed of each planned leg. According to the invention, the ship speed can be predicted and optimized.

Description

Technical Field

[0001] The present invention belongs to the technical field of ship operation management, and specifically relates to a ship speed prediction and optimization method, device and readable storage medium. Background Art

[0002] While the shipping industry is driving global economic development, its greenhouse gas emissions are also rising year by year. Ship speed is a key parameter for calculating fuel consumption and emissions, and is an important factor in evaluating the effectiveness of the Energy Efficiency Design Index (EEDI) and the Ship Energy Efficiency Management Plan (SEEMP). In order to improve fuel efficiency and reduce greenhouse gas emissions, it is necessary to accurately predict ship speed.

[0003] In actual navigation, it is difficult for ships to sail at the planned speed due to the complexity of the ships themselves and the influence of the resistance of the marine environment. Existing ship speed prediction technologies, such as ship model tests and fluid dynamics calculation methods, not only consume a lot of time and money, but also have difficulty in accurately exploring the relationship between multiple factors, resulting in failure to achieve ideal results. At the same time, the existing ship speed optimization technology cannot accurately predict the actual speed of the ship in a changing marine environment, which also leads to failure to achieve ideal results. Speed ​​prediction and optimization of planned voyages is one of the key factors in optimizing ship operations and is crucial to improving ship operating efficiency.

[0004] In summary, existing ship speed prediction technologies are mostly based on physical models established by theoretical derivation and ship model experiments, which are difficult to accurately describe the impact of complex marine environments on ship conditions, and many physical models ignore relevant features such as ocean currents to simplify the models. Therefore, predicting ship speed based on physical models is not accurate. Summary of the invention

[0005] In view of the above defects or improvement needs of the prior art, the present invention provides a ship speed prediction and optimization method, device and readable storage medium, which solves the problem of inaccurate ship speed prediction based on a data-driven method of machine learning. Compared with the physical model, the data-driven method based on machine learning can avoid the subjective assignment of parameters, and can comprehensively consider more factors affecting the ship's speed, and explore the nonlinear relationship between multiple factors, so as to establish a more accurate nonlinear ship speed prediction model. The accuracy of the ship speed prediction model is further improved by tuning the hyperparameters of the machine learning model through the particle swarm algorithm. The speed optimization model based on the high-accuracy ship speed prediction makes the predicted ship speed more close to the actual speed of the ship, thereby reducing the impact of meteorological information on the ship's speed, thereby improving the accuracy of the calculation of the ship's navigation cost, and better realizing the optimization of the ship's speed.

[0006] In a first aspect of the present invention, a method for predicting and optimizing ship speed is provided, the method comprising:

[0007] A ship speed prediction model is established with main engine power, ship heading, current velocity, current direction relative to the ship heading, wind speed, wind direction relative to the ship heading, effective wave height, wave direction relative to the ship heading, and combined period of wind waves and swells as inputs and ship speed as output;

[0008] Obtain ship route planning information and marine meteorological forecast information, including current velocity, current direction, wind speed, wind direction, significant wave height, wave direction, and the combined period of wind waves and swells;

[0009] Obtaining a planned ship route according to the ship route planning information, and dividing the planned ship route into a number of planned segments;

[0010] Obtain the ship's heading for each planned route, and obtain relevant information for each planned route in combination with marine meteorological forecast information, including ship's heading, current speed, current direction relative to the ship's heading, wind speed, wind direction relative to the ship's heading, significant wave height, wave direction relative to the ship's heading, and combined period of wind waves and swells;

[0011] According to the relevant information of different main engine powers and each planned section, combined with the ship speed prediction model, the actual ship speed of each planned section of the ship under the corresponding meteorological and sea conditions and different main engine powers is predicted;

[0012] A cost function is established based on the main engine power and the corresponding actual ship speed, and a ship speed optimization model is constructed in combination with constraint conditions;

[0013] Based on the ship speed prediction model and the ship speed optimization model, the optimized speed of each planned section is obtained.

[0014] Furthermore, a cost function is established according to the main engine power and the corresponding actual ship speed, and a ship speed optimization model is constructed in combination with constraint conditions, including:

[0015] The total cost of the voyage is calculated based on the following formula:

[0016]

[0017] Among them, cost is the cost of the planned segment i, n is the total number of planned segments; v0 is the designed speed, v i is the ideal speed under a given main engine power, h0 represents the fuel consumption rate under the design speed; D i represents the distance of the i-th planned flight segment; v iris the actual ship speed at the same main engine power in the i-th planned section, which is predicted by the ship speed prediction model; a1 and a2 are coefficients, corresponding to the fuel cost and time cost respectively;

[0018] At the same time, the following constraints are met:

[0019] T n ≤T max , ensuring that the ship arrives at the destination port on time and complies with the specified latest arrival time; among them, T i It represents the total time of the first i planned flight segments, T0 is 0;

[0020] Set the minimum and maximum limits of the ideal speed of the ship in each planned section; among them, and are the minimum speed and maximum speed respectively;

[0021] T i ≥0, requiring the start time of each planned flight segment to be non-negative to ensure the feasibility of the flight plan.

[0022] Furthermore, based on the ship speed prediction model and the ship speed optimization model, the optimized speed of each planned section is obtained, including:

[0023] For each planned route, the actual ship speed of the ship under different main engine powers is obtained, so as to obtain the cost of each planned route under different actual ship speeds;

[0024] Combined with the constraints, the actual ship speed of each planned section corresponding to the lowest cost of the total voyage is obtained, and then the ship is controlled to sail at the corresponding actual ship speed in the corresponding planned section.

[0025] Furthermore, a ship speed prediction model based on particle swarm optimization algorithm is established, including:

[0026] The ship speed prediction model is the Catboost model, and the hyperparameters of the Catboost model are optimized based on the particle swarm algorithm; the hyperparameters of the Catboost model include the maximum number of tree iterations, tree depth, and learning rate learning_rate;

[0027] Determine the dimension according to the number of parameters to be optimized, and then randomly initialize the position and velocity of the particles;

[0028] The performance of the ship speed prediction model is used as the fitness value, and the optimal Catboost model hyperparameters are obtained through iteration, thereby obtaining a ship speed prediction model based on the particle swarm optimization algorithm.

[0029] Furthermore, the planned ship route is divided into several planned segments, including:

[0030] According to the time resolution of the marine meteorological forecast information and the designed speed, planned waypoints are inserted on the planned ship route, and the planned ship route is divided into multiple pre-planned segments;

[0031] According to the changes in the ship's course in the pre-planned segment and the adjacent pre-planned segments and the changes in the marine meteorological forecast information, the pre-planned segment is further divided or combined to obtain the final planned segment.

[0032] Furthermore, a ship speed prediction model is established, including:

[0033] Obtain ship navigation history information and ocean meteorological history information;

[0034] Based on the method of single-source data cleaning and multi-source data fusion, a multi-source data aggregation processing model is constructed, and single-source data cleaning and multi-source data fusion are performed on the acquired ship navigation history information and marine meteorological history information to obtain a feature set;

[0035] Based on the Pearson correlation analysis method, the key feature set that has a significant impact on the ship speed is screened out from the feature set, the key features are used as the input of the ship speed prediction model, and the ship speed prediction model is trained using the key feature set;

[0036] Among them, obtaining ship navigation history information and ocean meteorological history information includes:

[0037] Obtain ship AIS data, including time, location, and corresponding ship speed, main engine power, and ship heading; where the location includes longitude and latitude;

[0038] Obtain historical marine meteorological information, including time, location, and corresponding current velocity, current direction, wind speed, wind direction, significant wave height, wave direction, and the combined period of wind waves and swells.

[0039] Furthermore, single-source data cleaning includes:

[0040] For each single-source data, outliers are identified and replaced with missing values, and then the single-source data are filled based on the tensor decomposition method, and the filled data are standardized using the Z-score method;

[0041] Among them, the identification methods include: for significant abnormal data, multidimensional clustering and fixed threshold method are used for identification; for non-significant abnormal data, theoretical knowledge and empirical models are used for identification;

[0042] For the ship speed and main engine power, two-dimensional tensor filling is used as follows:

[0043]

[0044] Among them, q MT is the ship speed or main engine power at the Mth position of the ship at time T; d M is the Mth position, t T is the Tth time; U and V are column vectors and row vectors of two dimensions M and T, and u and v are element values;

[0045] For other navigation parameters except ship speed and main engine power, the four-dimensional space is composed of ship speed, position, time and other navigation parameters, and then filled based on the four-dimensional tensor decomposition method.

[0046] Furthermore, multi-source data fusion includes:

[0047] Different data sets are aligned based on spatiotemporal attribute matching, and features with the same attributes are judged and converted to achieve consistent expression of features. Then, a data similarity measurement matrix is ​​constructed by combining the hash function method and Euclidean distance to identify redundant attributes, and data redundancy is eliminated through attribute reduction.

[0048] If there are differences in data records with the same attributes in heterogeneous data sets, the distribution statistics method is used to define the rationality of data from different sources, and the mean of the reasonable data is taken for numerical merging to achieve strong coupling integration of multi-source data.

[0049] According to a second aspect of the present invention, there is provided a ship speed prediction and optimization device, comprising: a processor and a memory, the memory storing programs or instructions executable on the processor, the programs or instructions implementing the steps of any one of the above-mentioned ship speed prediction and optimization methods when executed by the processor.

[0050] According to a third aspect of the present invention, there is provided a readable storage medium having a program or instruction stored thereon, which, when executed by a processor, implements the steps of any one of the above-mentioned ship speed prediction and optimization methods.

[0051] In general, the above technical solutions conceived by the present invention can achieve the following beneficial effects compared with the prior art:

[0052] The present invention provides a ship speed prediction and optimization method, device and readable storage medium, constructs a ship speed prediction model and a ship speed optimization model, and by considering the impact of the marine meteorological environment on the ship when sailing in the ocean and the characteristics of the ship itself, it can quickly and at a low cost predict the actual speed that is more appropriate for the ship's navigation, so as to solve the defects of the existing ship speed prediction technology that not only consumes a lot of time and money, but also is difficult to accurately explore the relationship between the influence of multiple factors. In addition, through accurate and fast ship speed prediction, the impact of inaccurate speed on cost calculation in speed optimization is reduced, the accuracy of calculation of ship navigation costs is improved, and a cost function is established according to the main engine power and the actual ship speed, and a ship speed optimization model is constructed in combination with constraint conditions, so as to solve the defect that the existing ship speed optimization cannot achieve ideal results.

[0053] In addition, the present invention constructs a multi-source data aggregation processing model based on a unique single-source data cleaning and multi-source data fusion strategy, and trains a machine learning model based on particle swarm optimization through the fused data set, thereby constructing a ship speed prediction model, thereby improving the prediction performance of the ship speed prediction model. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 It is a flow chart of the ship speed prediction and optimization method;

[0055] Figure 2 This is a flow chart of optimizing hyperparameters of machine learning models based on particle swarm algorithm;

[0056] Figure 3 A schematic diagram of a process for optimizing ship speed based on a ship speed prediction model and a ship speed optimization model;

[0057] Figure 4 This is a schematic diagram of the structure of the ship speed prediction and optimization device;

[0058] Figure 5 This is a schematic diagram of the speed optimization results under different wind and wave conditions;

[0059] Figure 6 Schematic diagram of the cost of each leg using the optimized speed. DETAILED DESCRIPTION

[0060] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention. In addition, the technical features involved in the various embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0061] Existing ship speed prediction technologies are mostly based on physical models established by theoretical derivation and ship model experiments. It is difficult to accurately describe the impact of complex marine environment on ship status, and many physical models ignore relevant features such as ocean currents to simplify the model. Therefore, the prediction of ship speed based on physical models is not accurate. Compared with physical models, the data-driven method based on machine learning can avoid subjective parameter assignment, and can comprehensively consider more factors affecting ship speed, and explore the nonlinear relationship between multiple factors, so as to establish a more accurate nonlinear ship speed prediction model. The hyperparameters of the machine learning model are tuned by particle swarm algorithm to further improve the accuracy of the ship speed prediction model. The speed optimization model based on high-accuracy ship speed prediction makes the predicted ship speed more close to the actual speed of the ship, thereby reducing the impact of meteorological information on the ship speed, and then improving the accuracy of the calculation of the ship's navigation cost. The cost function is established according to the main engine power and the actual ship speed, and the ship speed optimization model is constructed in combination with the constraint conditions, which can better achieve the optimization of the ship speed.

[0062] like Figure 1 As shown, according to one embodiment of the present invention, a method for predicting and optimizing the speed of an ocean-going ship based on multi-source heterogeneous sensor data in a variable marine environment is proposed, comprising the following steps:

[0063] S101, obtaining ship navigation history information and ocean meteorological history information;

[0064] S102, constructing a multi-source data aggregation processing model based on a single-source data cleaning and multi-source data fusion method to process the acquired ship navigation history information and marine meteorological history information;

[0065] S103, establishing a ship speed prediction model based on a particle swarm optimization algorithm according to the processed ship navigation history information and ocean meteorological history information;

[0066] S104, obtaining ship route planning information and marine weather forecast information;

[0067] S105, constructing a ship speed optimization model according to the cost function and the constraint conditions;

[0068] S106, based on the speed prediction model and the speed optimization model, a ship speed prediction and optimization model is established to optimize the speed of the planned route.

[0069] The present invention aggregates ship navigation data and ocean meteorological data to make the predicted ship speed more close to the actual speed of the ship, thereby reducing the impact of meteorological information on the ship speed, thereby improving the accuracy of the calculation of the ship navigation cost and better realizing the optimization of the ship speed.

[0070] Specifically, first, obtain the ship's navigation history information and marine meteorological history information. For example, the time, longitude and latitude coordinates, heading, speed, main engine power, draft and other ship navigation information during the actual voyage of the ship, as well as the meteorological information around the ship. By establishing a multi-source data aggregation processing model, the ship's navigation information and meteorological information are processed and integrated. Then, the machine learning model based on particle swarm optimization is trained through the fused data set to build a ship speed prediction model. Secondly, obtain the ship's route planning information and marine meteorological forecast information, and process the fusion of the ship's planned waypoints and meteorological forecast information through the multi-source data aggregation processing model. Among them, the planned waypoints are set according to the design speed of the ship under ideal circumstances. The waypoints are divided according to the marine meteorological conditions of the planned waypoints, and the speed of each section is optimized according to the cost function and its limiting conditions to achieve speed optimization.

[0071] In some embodiments, obtaining ship navigation history information and ocean meteorological history information includes:

[0072] The data set M1 is established by obtaining ship AIS data and marine meteorological historical information from the weather forecast center; the ship AIS data includes ship speed, heading, time, longitude, latitude, main engine power, ship static data (such as ship length, ship width, draft, etc.). The marine meteorological historical information includes current velocity (m / s), current direction (°), wind speed (m / s), wind direction (°), significant wave height (m), wave direction (°), and the combined period of wind waves and swells (seconds).

[0073] In some embodiments, a multi-source data aggregation processing model is constructed based on a method of single-source data cleaning and multi-source data fusion, including:

[0074] Identify and process data based on historical ship navigation data and marine meteorological information;

[0075] For missing, outlier and other significant abnormal data, multidimensional clustering and fixed threshold methods are used for identification; for non-significant abnormal data generated in special maneuvering modes such as "ramming", theoretical knowledge and empirical models are used for identification; then the missing values ​​are used to replace the abnormal values, and then the single-source data set is filled based on the tensor decomposition method, and the filled data is standardized using the Z-score method, finally achieving high-quality cleaning of the original data. The ship speed and main engine power data belong to a two-dimensional time series, and a two-dimensional matrix is ​​constructed; for other navigation parameters of the ship, a four-dimensional space is formed by speed, position (longitude and latitude), time and other navigation parameters. Take the two-dimensional tensor filling as an example, where d represents the different position points of the ship and t represents the time series.

[0076] Therefore, a two-dimensional matrix (such as the speed of a ship at different time points) can be decomposed into:

[0077]

[0078] Among them, q MT is the ship attribute value (such as speed, main engine power) at the Mth point at time T; U and V are column vectors and row vectors with two dimensions of M and T. Decomposition can reveal hidden patterns in time and space and be used to predict and fill missing data points to achieve single-source data cleaning.

[0079] Assume that the constructed two-dimensional matrix Q = UV, U is a matrix representing the spatial dimension (position point or parameter), and V is a matrix representing the time dimension. By decomposing the known data, the correlation in time and space is found. For example, the change pattern between adjacent time points and adjacent spatial positions can be reflected by the decomposed matrices U and V. The pattern reflects the changing trend of the speed or other parameters of the ship under certain specific conditions during the voyage. Use the decomposed matrices U and V to fill the missing values ​​in the original matrix Q. For example, the ship speed data at certain time points is missing, and the ship speed at the missing time point can be inferred from the patterns of the previous and next time points. Similarly, for missing data in the spatial dimension (such as the missing main engine power at certain position points), it can be reconstructed through the decomposed spatial pattern.

[0080] It should be noted that navigation parameters refer to other characteristic variables related to ship speed, such as ship heading, wind speed, and wind direction. Other navigation parameters of ships refer to other characteristic data except time, position, ship speed, and main engine power. By integrating various navigation parameters at different times and positions, an overall navigation trajectory can be formed, and trends and correlations in time and space can be analyzed. For example, how wind speed and wind direction change with time and position, and how they affect ship speed. It also better reveals abnormal points in the data, especially for some parameters with strong temporal and spatial correlations.

[0081] Different data sets are aligned based on spatiotemporal attribute matching, and the same attribute features (such as wind direction and ship relative wind direction, ship navigation history data may also contain ship relative wind direction data) are judged and converted (the direction of meteorological data is converted according to the ship's heading, and converted to the direction relative to the heading) to achieve consistent expression of features. Afterwards, the data similarity measurement matrix is ​​constructed by combining the hash function method and Euclidean distance to identify redundant attributes, and data redundancy is achieved through attribute reduction. If there are differences in data records with the same attributes in heterogeneous data sets, the distribution statistics method is used to define the rationality of data from different sources, and the mean of reasonable data is taken for numerical merging (average processing is performed on multiple different values ​​of the same attribute), and finally a multi-source data aggregation processing model is constructed to achieve strong coupling integration of multi-source data.

[0082] By performing single-source data cleaning and multi-source data fusion on ship navigation data and meteorological data, a multi-source data aggregation processing model is constructed. Based on the multi-source data aggregation processing model and data set M1, the external marine meteorological data and ship AIS data are cleaned and fused to establish data set S1.

[0083] In some embodiments, a speed prediction model based on a particle swarm optimization algorithm is established according to the ship's navigation history information and the ocean meteorological history information, including:

[0084] Based on the Pearson correlation analysis method, the key feature set that has a significant impact on the ship speed is screened out;

[0085] Optimize machine learning model hyperparameters based on particle swarm algorithm;

[0086] A speed prediction model is established by using the data set S1 to train the optimized machine learning model.

[0087] Specifically, firstly, the Pearson correlation analysis method was used to calculate the correlation coefficients between different features and construct a correlation coefficient matrix, and the key feature set that has a significant impact on the ship speed was screened out. Finally, nine input variables were selected from the data set: main engine power, ship heading (°), flow velocity (m), relative flow direction to the ship heading (°), wind speed (m / s), relative wind direction (°), effective wave height (m), relative wave direction (°) and combined period of wind waves and swells (seconds), and the ship speed (knot) was selected as the output variable, and the data set S1 was updated.

[0088] Secondly, if Figure 2 As shown in the figure, the hyperparameters of the machine learning model are optimized based on the particle swarm algorithm, including:

[0089] Determine the dimension N according to the number of parameters n that need to be optimized, and then randomly initialize the position and velocity of the particles;

[0090] The position attribute of each particle is an N-dimensional vector, whose range refers to the entire search space. The components of each dimension correspond to different parameters of the machine learning model, and the initialization range of each dimension is different;

[0091] The position vector of the i-th particle at time t can be expressed as:

[0092]

[0093] Since all particles move in the same search space, when t = 0, the velocity can be initialized to the (0,1) range of each dimension;

[0094] The velocity vector of the i-th particle at time t can be expressed as:

[0095]

[0096] Assign the position vector to the corresponding parameter of the model, take the performance on the training set as the initial fitness value, set the mean square error (MSE) as the target optimization function of PSO, and the fitness value of the i-th particle at time t is:

[0097] F i(t) =(P i(t) →Model| traningset ) [metric＝MSE]

[0098] For the i-th particle, its individual optimal value at time t can be expressed as:

[0099] Pbest i(t) =max(F i(j) ),0≤j≤t

[0100] Assuming there are m particles, the global optimum at time t can be expressed as:

[0101] Gbest (t) =max(Pbest k(t) ),1≤k≤m

[0102] According to the target error value or the maximum number of iterations, the model training is terminated to obtain the optimized parameter values ​​of the machine learning model.

[0103] In this embodiment, the model is a Catboost model. Figure 2 As shown, the hyperparameters of the Catboost model are optimized based on the particle swarm algorithm. The dimension is determined to be 3 according to the number of parameters to be optimized (iterations, depth, learning_rate), and then the position and speed of the particles are randomly initialized;

[0104] The position attribute of each particle is a 3D vector whose range refers to the entire search space. The components of each dimension correspond to different parameters of the Catboost model, and the initialization range of each dimension is different.

[0105] The position vector of the i-th particle at time t can be expressed as:

[0106]

[0107] Since all particles move in the same search space, when t = 0, the velocity can be initialized to the (0,1) range of each dimension;

[0108] The velocity vector of the i-th particle at time t can be expressed as:

[0109]

[0110] Assign the position vector to the corresponding parameters of the model, use the performance on the training set as the initial fitness value, set the mean square error as the target optimization function of PSO, and the fitness value of the i-th particle at time t is:

[0111] F i(t) =(P i(t) →Model| traningset ) [metric＝MSE]

[0112] For the i-th particle, its individual optimal value at time t can be expressed as:

[0113] Pbest i(t) =max(F i(j) ),0≤j≤t

[0114] Assuming there are m particles, the global optimum at time t can be expressed as:

[0115] Gbest (t) =max(Pbest k(t) ),1≤k≤m

[0116] According to the target error value or the maximum number of iterations, the model training is terminated to obtain the optimized parameter values ​​of the machine learning model, as shown in Table 1 below.

[0117] Table 1 Optimization parameter values

[0118]

[0119] The speed prediction model was established by using the machine learning model optimized by training the dataset S1, and four standard regression performance indicators were used to evaluate the performance of the ship speed prediction model, including mean square error, root mean square error (RMSE), mean absolute error (MAE) and mean absolute percentage error (MAPE).

[0120] In the reference example, 80% of the data (6952 items) are randomly selected from the data set as the training set to build the ship speed prediction model, and the remaining 20% ​​(1738 items) are used as the test set. The performance of the trained PSO-Catboost and Catboost speed prediction models is shown in Table 2.

[0121] Table 2 Performance of speed prediction model

[0122]

[0123] The PSO-Catboost speed prediction model performs better than the Catboost speed prediction model in all four evaluation indicators. It can be concluded that optimizing machine learning hyperparameters based on particle swarm algorithm can effectively improve the prediction performance of machine learning models.

[0124] In some embodiments, obtaining ship route planning information and marine weather forecast information includes:

[0125] The ship route planning information including the departure port, destination port, and estimated arrival time and the marine meteorological forecast information generated by the Global Ensemble Forecast System from the National Oceanic and Atmospheric Administration are obtained to establish the data set M2.

[0126] In some embodiments, a speed optimization model is constructed based on the cost function and the constraint conditions, including:

[0127] Establish route cost functions and constraints;

[0128] Based on the formula:

[0129]

[0130] Calculate the cost of the voyage, where v0 is the design speed (16 knots in this example), v i is the ideal speed (no wind and no waves) under a given power setting (i.e. main engine power), v ir is the actual speed under the same power, D i represents the distance of the i-th segment, h0 represents the fuel consumption rate at the design speed, and a1 and a2 correspond to the fuel cost and time cost respectively; divided by 24 represents the conversion to days. The variable i represents the index of the segment on the scheduled route, and i∈I means that i is an element in the segment set I.

[0131] At the same time, the constraints are met: To ensure the integrity of the time sequence, T n ≤T max Ensure that the vessel arrives at the destination port on time and adheres to the specified latest arrival time. Set the minimum and maximum speed limits for the ship in each section. The starting time of each flight segment is required to be non-negative to ensure the feasibility of the flight plan.

[0132] The optimal decision for the current segment is found by calculating the cost of different speed options for the current segment and combining it with the optimal solution of the previous segment, until it recurses to the last segment to solve the optimal decision for each segment.

[0133] It should be noted that there is a corresponding relationship between the main engine power and the ideal speed, which can be obtained through simulation or actual measurement.

[0134] In some embodiments, based on the speed prediction model and the speed optimization model, a ship speed prediction and optimization model is established to optimize the speed of the route, such as Figure 3 As shown, including:

[0135] Step S1061, according to the time resolution of the marine meteorological forecast information M2, the ship speed is assumed to be sailing at the design speed under ideal conditions, and the planned waypoints are inserted;

[0136] Based on the multi-source data aggregation processing model, the planned waypoints and marine meteorological forecast data are aggregated to establish data set S2.

[0137] Exemplarily, the marine weather forecast information is updated every hour, and is multiplied by the design speed to obtain a distance interval, and then a planned waypoint is inserted at each distance interval along the planned route.

[0138] Step S1062, judging the marine meteorological conditions of each planned waypoint according to the marine meteorological data in the data set S2, and dividing the planned route into i planned waypoints and i-1 planned flight segments accordingly.

[0139] The planned segments can be further divided or combined according to the changes in the ship's course in the planned segment and the adjacent planned segments and the changes in the marine meteorological forecast information to obtain the final planned segment. For example, if the ship's course changes significantly in the planned segment, a planned waypoint is inserted. If the conditions of multiple planned segments are similar, the planned segments can be merged.

[0140] In the reference example, the route is divided into 19 segments according to the sea conditions and Beaufort wind scale between waypoints. The specific information is shown in Table 3 below.

[0141] Table 3 Flight segment table

[0142]

[0143]

[0144] Step S1063, according to the ocean meteorological conditions of each section, the actual speed is predicted based on the speed prediction model, and the optimal speed is selected based on the route optimization model to minimize the cost and ensure that the destination port is reached within the scheduled time.

[0145] Step S1064, establishing a ship speed prediction and optimization model to achieve speed optimization to meet the latest arrival time while maximizing the reduction of the total operating cost of the multi-segment voyage.

[0146] By adopting the above technical solutions, a ship speed prediction and optimization model is constructed.

[0147] According to a second aspect of the present invention, Figure 4 As shown in FIG. 1 , a structural block diagram of a ship speed prediction device 200 is proposed. Figure 4 As shown, the ship speed prediction and optimization device 200 mainly includes:

[0148] The first acquisition module 201 is used to acquire ship navigation information and ocean meteorological historical information;

[0149] Establishing module one 202, for constructing a multi-source data aggregation processing model based on a method of single-source data cleaning and multi-source data fusion;

[0150] Establishing module 203, for establishing a speed prediction model based on particle swarm optimization algorithm according to ship navigation information and marine meteorological historical information;

[0151] The second acquisition module 204 is used to acquire ship route planning information and marine weather forecast information;

[0152] Establishing module three 205, for constructing a speed optimization model according to the cost function and the constraint conditions;

[0153] The application module 1 206 is used to establish a ship speed prediction and optimization model based on the speed prediction model and the speed optimization model, and optimize the speed of the planned route.

[0154] As an optional implementation of this embodiment, the application module 1 202 includes:

[0155] Data processing module: used to identify abnormal data based on ship navigation data and marine meteorological information using multidimensional clustering and fixed threshold method, and interpolate and fill in;

[0156] Data fusion submodule: used to align different data sets based on spatiotemporal attribute matching, and to determine and convert features with the same attributes to achieve consistent expression of features.

[0157] As an optional implementation of this embodiment, the second module 203 is established, including:

[0158] Feature screening submodule: Based on the Pearson correlation analysis method, the key feature set that has a significant impact on the ship speed is screened out;

[0159] Hyperparameter optimization submodule: optimizes the hyperparameters of machine learning models based on the particle swarm optimization algorithm;

[0160] Model training submodule: Establish a speed prediction model based on the machine learning model optimized after training based on ship navigation data and marine meteorological information.

[0161] As an optional implementation of this embodiment, the establishment module 3 205 includes:

[0162] By establishing cost functions and constraints, the costs of different speed options for the current segment are calculated to solve the optimal decision for each segment and build a speed optimization model.

[0163] As an optional implementation of this embodiment, the application module 1 206 includes:

[0164] Application submodule 1: used to aggregate planned waypoints with marine meteorological forecast data based on a multi-source data aggregation processing model;

[0165] Application submodule 2: used to predict the actual speed based on the speed prediction model according to the marine meteorological conditions of each section;

[0166] Application submodule three: used to select the optimal speed based on the route optimization model decision according to the marine meteorological conditions of each section;

[0167] Establish submodule: used to establish ship speed prediction and optimization model, realize speed optimization to meet the latest arrival time while maximizing the total operating cost of multi-segment voyage.

[0168] According to a third aspect of the present invention, a device for predicting and optimizing the speed of an ocean-going ship based on multi-source heterogeneous sensor data in a changeable ocean environment is proposed, comprising: a processor and a memory, the memory storing programs or instructions that can be executed on the processor, and the programs or instructions, when executed by the processor, implement the steps of the method for predicting and optimizing the speed of an ocean-going ship based on multi-source heterogeneous sensor data in a changeable ocean environment as provided in the first aspect.

[0169] According to a fourth aspect of the present invention, a readable storage medium is proposed, on which a program or instruction is stored. When the program or instruction is executed by a processor, a method for predicting and optimizing the speed of an ocean-going ship based on multi-source heterogeneous sensor data in a changeable marine environment as described in any of the above technical solutions is implemented.

[0170] Figure 5 The optimized speed of a 6238TEU container ship under different wind and wave conditions is shown. Figure 6 As shown in the figure, the optimized ship speed significantly reduces the operating costs of container ships in each segment compared with the ship's design speed. This strategic approach saves approximately $81,011.16 in costs, accounting for approximately 4.6% of total operating expenses.

[0171] In summary, the ship speed prediction and optimization model proposed in the present invention performs better than existing methods in terms of both ship speed prediction and speed optimization. This invention mainly includes a multi-source data aggregation processing model based on a unique single-source data cleaning and multi-source data fusion strategy and a machine learning ship speed prediction and optimization model based on a particle swarm algorithm. This invention can quickly and at a lower cost predict the actual speed that is more appropriate for the ship's navigation by considering the impact of the marine meteorological environment on the ship when it is sailing in the ocean and the characteristics of the ship itself, so as to solve the problem that the existing ship speed prediction technology not only consumes a lot of time and money, but also has difficulty in accurately exploring the relationship between multiple factors that influence each other. Through accurate and rapid ship speed prediction, the impact of inaccurate speed on cost calculation in speed optimization is reduced, the accuracy of the calculation of ship navigation costs is improved, and the defect that the existing ship speed optimization cannot achieve ideal results is solved.

[0172] It should be noted that the size of the serial numbers of the steps in the above embodiments does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0173] It should be pointed out that, according to the needs of implementation, the various steps / components described in this application can be split into more steps / components, and two or more steps / components or partial operations of steps / components can be combined into new steps / components to achieve the purpose of the present invention.

[0174] It will be easily understood by those skilled in the art that the above are only preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for predicting and optimizing ship speed, characterized in that: The method includes: A ship speed prediction model is established with main engine power, ship heading, current velocity, current direction relative to the ship heading, wind speed, wind direction relative to the ship heading, effective wave height, wave direction relative to the ship heading, and combined period of wind waves and swells as inputs and ship speed as output; Obtain ship route planning information and marine meteorological forecast information, including current velocity, current direction, wind speed, wind direction, significant wave height, wave direction, and the combined period of wind waves and swells; Obtaining a planned ship route according to the ship route planning information, and dividing the planned ship route into a number of planned segments; Obtain the ship's heading for each planned route, and obtain relevant information for each planned route in combination with marine meteorological forecast information, including ship's heading, current speed, current direction relative to the ship's heading, wind speed, wind direction relative to the ship's heading, significant wave height, wave direction relative to the ship's heading, and combined period of wind waves and swells; According to the relevant information of different main engine powers and each planned section, combined with the ship speed prediction model, the actual ship speed of each planned section of the ship under the corresponding meteorological and sea conditions and different main engine powers is predicted; A cost function is established based on the main engine power and the corresponding actual ship speed, and a ship speed optimization model is constructed in combination with constraint conditions; Based on the ship speed prediction model and the ship speed optimization model, the optimized speed of each planned section is obtained.

2. The ship speed prediction and optimization method according to claim 1, characterized in that: The cost function is established according to the main engine power and the corresponding actual ship speed, and the ship speed optimization model is constructed in combination with the constraint conditions, including: The total cost of the voyage is calculated based on the following formula: Among them, cost is the cost of the planned segment i, n is the total number of planned segments; v0 is the designed speed, v i is the ideal speed under a given main engine power, h0 represents the fuel consumption rate under the design speed; D i represents the distance of the i-th planned flight segment; v ir is the actual ship speed at the same main engine power in the i-th planned section, which is predicted by the ship speed prediction model; a1 and a2 are coefficients, corresponding to the fuel cost and time cost respectively; At the same time, the following constraints are met: T n ≤T max , ensuring that the ship arrives at the destination port on time and complies with the specified latest arrival time; among them, T i It represents the total time of the first i planned flight segments, T0 is 0; Set the minimum and maximum limits of the ideal speed of the ship in each planned section; among them, and are the minimum speed and maximum speed respectively; T i ≥0, requiring the start time of each planned flight segment to be non-negative to ensure the feasibility of the flight plan.

3. The ship speed prediction and optimization method according to claim 2 is characterized in that: Based on the ship speed prediction model and the ship speed optimization model, the optimized speed of each planned section is obtained, including: For each planned route, the actual ship speed of the ship under different main engine powers is obtained, so as to obtain the cost of each planned route under different actual ship speeds; Combined with the constraints, the actual ship speed of each planned section corresponding to the lowest cost of the total voyage is obtained, and then the ship is controlled to sail at the corresponding actual ship speed in the corresponding planned section.

4. The ship speed prediction and optimization method according to claim 1, characterized in that: A ship speed prediction model based on particle swarm optimization algorithm is established, including: The ship speed prediction model is the Catboost model, and the hyperparameters of the Catboost model are optimized based on the particle swarm algorithm; the hyperparameters of the Catboost model include the maximum number of tree iterations, tree depth, and learning rate learning_rate; Determine the dimension according to the number of parameters to be optimized, and then randomly initialize the position and velocity of the particles; The performance of the ship speed prediction model is used as the fitness value, and the optimal Catboost model hyperparameters are obtained through iteration, thereby obtaining a ship speed prediction model based on the particle swarm optimization algorithm.

5. The ship speed prediction and optimization method according to claim 1, characterized in that: Divide the planned ship route into several planned segments, including: According to the time resolution of the marine meteorological forecast information and the designed speed, planned waypoints are inserted on the planned ship route, and the planned ship route is divided into multiple pre-planned segments; According to the changes in the ship's course in the pre-planned segment and the adjacent pre-planned segments and the changes in the marine meteorological forecast information, the pre-planned segment is further divided or combined to obtain the final planned segment.

6. The ship speed prediction and optimization method according to claim 1, characterized in that: Establish a ship speed prediction model, including: Obtain ship navigation history information and ocean meteorological history information; Based on the method of single-source data cleaning and multi-source data fusion, a multi-source data aggregation processing model is constructed, and single-source data cleaning and multi-source data fusion are performed on the acquired ship navigation history information and marine meteorological history information to obtain a feature set; Based on the Pearson correlation analysis method, the key feature set that has a significant impact on the ship speed is screened out from the feature set, the key features are used as the input of the ship speed prediction model, and the ship speed prediction model is trained using the key feature set; Among them, obtaining ship navigation history information and ocean meteorological history information includes: Obtain ship AIS data, including time, location, and corresponding ship speed, main engine power, and ship heading; where the location includes longitude and latitude; Obtain historical marine meteorological information, including time, location, and corresponding current velocity, current direction, wind speed, wind direction, significant wave height, wave direction, and the combined period of wind waves and swells.

7. The ship speed prediction and optimization method according to claim 6, characterized in that: Single-source data cleaning includes: For each single-source data, outliers are identified and replaced with missing values, and then the single-source data are filled based on the tensor decomposition method, and the filled data are standardized using the Z-score method; Among them, the identification methods include: for significant abnormal data, multidimensional clustering and fixed threshold method are used for identification; for non-significant abnormal data, theoretical knowledge and empirical models are used for identification; For the ship speed and main engine power, two-dimensional tensor filling is used as follows: Among them, q MT is the ship speed or main engine power at the Mth position of the ship at time T; d M is the Mth position, t T is the Tth time; U and V are column vectors and row vectors of two dimensions M and T, and u and v are element values; For other navigation parameters except ship speed and main engine power, the four-dimensional space is composed of ship speed, position, time and other navigation parameters, and then filled based on the four-dimensional tensor decomposition method.

8. The ship speed prediction and optimization method according to claim 6, characterized in that: Multi-source data fusion includes: Different data sets are aligned based on spatiotemporal attribute matching, and features with the same attributes are judged and converted to achieve consistent expression of features. Then, a data similarity measurement matrix is ​​constructed by combining the hash function method and Euclidean distance to identify redundant attributes, and data redundancy is eliminated through attribute reduction. If there are differences in data records with the same attributes in heterogeneous data sets, the distribution statistics method is used to define the rationality of data from different sources, and the mean of the reasonable data is taken for numerical merging to achieve strong coupling integration of multi-source data.

9. A ship speed prediction and optimization device, characterized in that: include: A processor and a memory, wherein the memory stores programs or instructions that can be run on the processor, and when the programs or instructions are executed by the processor, the steps of the ship speed prediction and optimization method described in any one of claims 1 to 8 are implemented.

10. A readable storage medium, characterized in that: Programs or instructions are stored thereon, and when the programs or instructions are executed by the processor, the steps of the ship speed prediction and optimization method described in any one of claims 1 to 8 are implemented.

Images