A Ship Navigation Speed Optimization Control Method and Device Based on Multi-Objective Optimization

Abstract

The present application discloses a method and device for optimizing the navigation speed of a ship based on multi-objective optimization, belonging to the field of optimizing ship navigation control. The present application first establishes a simulation model of the ship's speed and power output, corrects the simulation results through weather forecast data, and further conducts computational analysis through the simulation model and multi-objective optimization algorithm to minimize the time and energy consumption during the ship's navigation, generating a power output control scheme, thereby controlling the ship's navigation speed by controlling the power output. Compared with the prior art, the present application significantly improves the ship's navigation efficiency and reduces energy costs by optimizing the control strategies of power output and navigation speed.

Description

Technical Field

[0001] This application relates to the field of ship navigation control optimization, and particularly to a ship navigation speed optimization control method and device based on multi-objective optimization. Background Art

[0002] Ship navigation is easily interfered by environmental factors such as wind waves, ocean currents, and weather. With the improvement of the accuracy of weather forecast data, by using accurate weather forecast data to optimize the speed control plan during navigation, the reaction speed of the crew to emergencies can be effectively improved, thereby enhancing the safety of navigation.

[0003] In the prior art, usually, a physical process simulation model is constructed, and the weather forecast model is embedded into the physical process simulation model to predict the ship speed under different environments and different power outputs. At the same time, a large number of sensors are set to obtain more accurate real-time weather data to ensure the accuracy of the physical process simulation model. However, the physical process simulation model used in this method has a large number of parameters, requires a large amount of observation data to adjust the parameters of the physical process simulation model, and has a high hardware cost, making it difficult to popularize for different ship types. At the same time, when estimating the influence of weather environment factors on the ship speed, the physical process simulation model mainly focuses on the instantaneous change influence and fails to consider the influence of persistent weather factors on the ship speed, resulting in higher energy consumption and longer time consumption in the subsequent optimized solutions.

[0004] Therefore, how to more accurately integrate weather forecast data into the ship speed optimization problem while ensuring the computational complexity, so as to reduce the navigation time and navigation energy consumption of the ship speed control scheme, is a technical problem that needs to be solved currently. Summary of the Invention

[0005] This application provides a ship navigation speed optimization control method based on multi-objective optimization, which can solve the technical problem that in the prior art, how to more accurately integrate weather forecast data into the ship speed optimization problem while ensuring the computational complexity, so as to reduce the navigation time and navigation energy consumption of the ship speed control scheme.

[0006] A ship navigation speed optimization control method based on multi-objective optimization provided by an embodiment of this application includes:

[0007] Obtain the navigation route data to be planned and weather forecast data; the navigation route data includes several stopping points, the optimal navigation routes between each pair of stopping points, and several preset speed change points corresponding to each of the optimal navigation routes; the preset speed change points are set according to the turning points of the corresponding optimal navigation routes.

[0008] Take the arrival order of each of the said stopping points as the first decision variable, and take the power output of each of the said preset variable-speed points corresponding to each of the said optimal navigation routes as the second decision variable;

[0009] According to the first decision variable, the second decision variable, and a preset power simulation model, determine the ship speeds corresponding to a number of preset variable-speed points, and correct the ship speeds through a preset long short-term memory network model in combination with the said weather forecast data; the power simulation model is a mapping relationship between power output and ship speed established under the condition of no environmental disturbance;

[0010] According to the first decision variable, the second decision variable, and the corrected ship speeds, respectively determine the first navigation time and the first navigation energy consumption, and take the first navigation time and the first navigation energy consumption as the first optimization goal and the second optimization goal respectively;

[0011] According to the first optimization goal and the second optimization goal, construct a multi-objective optimization problem, and solve the multi-objective optimization problem through a multi-objective optimization algorithm to obtain an optimal power output plan.

[0012] Compared with the prior art, the above embodiments have the following beneficial effects: Whenever the ship sails to the turning point of the optimal navigation route, it is easier for the influence of the weather environment on the ship speed to change at this time. Therefore, by setting the turning points of each optimal navigation route as preset variable-speed points, the sensitivity of subsequent ship speed optimization to weather forecast data is improved; further, through a simplified power simulation model, while solving the problem of low versatility caused by the overly complex physical process model, a long short-term memory network model that can capture the change characteristics of time-series data is introduced to further correct the ship speed. The above process of estimating the ship speed through the weather environment not only reduces the computational complexity, but also ensures the accuracy of ship speed estimation; finally, by establishing a multi-objective optimization problem and solving the optimal power output plan, the final power output plan can reduce both the navigation time and the navigation energy consumption during the navigation process.

[0013] Further, the step of determining the ship speeds corresponding to a number of preset variable-speed points according to the first decision variable, the second decision variable, and a preset power simulation model includes:

[0014] According to the first decision variable, determine the arrival order of each of the said stopping points;

[0015] According to the arrival order of each of the said stopping points, determine a number of first optimal navigation routes;

[0016] Take the preset variable-speed points corresponding to the first optimal navigation routes as the first preset variable-speed points;

[0017] Through the simulation model, the second decision variables corresponding to the first preset speed change points are mapped to corresponding speeds.

[0018] Compared with the prior art, the above embodiments have the following beneficial effects: In order to avoid the problem of excessive computational complexity in the optimization process caused by too many power output decisions corresponding to all preset speed change points in the entire solution space, the optimization space of the subsequent second decision variables is limited by the solution space defined by the first decision variable, thereby effectively improving the optimization speed; further, the second decision variables corresponding to the first preset speed change points are mapped through the simulation model, so that the sailing speed of the ship at each first preset speed change point can be preliminarily determined.

[0019] Further, the method for correcting the speed by combining the preset long short-term memory network model with the weather forecast data includes:

[0020] Obtain the first weather forecast data of each of the first optimal sailing routes within a preset time interval from the weather forecast data;

[0021] Determine the sailing order of each of the first optimal sailing routes according to the first decision variable;

[0022] In the sailing order, sequentially correct the speeds corresponding to the first preset speed change points by combining the preset long short-term memory network model with the first weather forecast data.

[0023] Compared with the prior art, the above embodiments have the following beneficial effects: Since the weather forecast has a time attribute, except for the initial departure time, the time when the ship reaches each first preset speed change point subsequently cannot be predicted in advance. Therefore, after determining several first optimal sailing routes, it is necessary to further consider the passing order of each first optimal sailing route. Starting from the first first preset speed change point (i.e., the initial starting point), combined with the first weather forecast data at the initial time, gradually iterate to determine the time period of the first weather forecast data required to correct the next first preset speed change point, thereby improving the accuracy of speed correction.

[0024] Further, the step of sequentially correcting the speeds corresponding to the first preset speed change points by combining the preset long short-term memory network model with the first weather forecast data in the sailing order includes:

[0025] Take the speed to be corrected currently as the first speed, and take the time point when the ship reaches the first preset speed change point corresponding to the first speed as the first time point, and determine whether the first preset speed change point corresponding to the first speed is the first preset speed change point corresponding to any of the first optimal sailing routes;

[0026] If so, the first sailing speed is corrected by the data-driven method in combination with the first weather forecast data at the first time point.

[0027] Otherwise, the time point when the ship reaches the first preset variable speed point corresponding to the last corrected sailing speed is used as the second time point, and the first sailing speed is corrected by the preset long short-term memory network model according to the first weather forecast data during the period from the first time point to the second time point.

[0028] Compared with the prior art, the above embodiments have the following beneficial effects: Since the ship starts from each docking point at the first first preset variable speed point of each first optimal sailing route, there are no accumulated weather environment factors affecting before the first first preset variable speed point. Therefore, the initial first sailing speed needs to be corrected with the first weather forecast data in the form of a time point. When the ship leaves the first first preset variable speed point, it begins to be affected by the accumulated weather environment factors. Therefore, according to the first weather forecast data experienced between the first preset variable speed points of the two vectors (at this time, the data is time series data within a time period), the characteristic information in the time series data is obtained through the preset long short-term memory network model to correct the first sailing speed, thereby improving the correction accuracy.

[0029] Further, the hidden state recurrence process of the preset long short-term memory network model includes:

[0030] Wherein, the preset long short-term memory network model includes: spatio-temporal feature gating, environmental disturbance gating, short-term memory unit and long-term memory unit;

[0031] Through the spatio-temporal feature gating, spatio-temporal features are extracted from the first weather forecast data at the current time step, and the short-term memory unit is updated through the spatio-temporal features.

[0032] Fuse the characteristic information of the long-term memory unit at the previous time step and the characteristic information of the short-term memory unit at the current time step to obtain the characteristic information of the long-term memory unit at the current time step.

[0033] Through the environmental disturbance gating, the disturbance weight is obtained in combination with the historical mean of the weather forecast.

[0034] According to the disturbance weight and the characteristic information of the short-term memory unit and the long-term memory unit at the current time step, the memory cell of the preset long short-term memory network model is updated.

[0035] According to the updated memory cell, the hidden state output by the preset long short-term memory network model at the current time step is obtained.

[0036] Compared with the prior art, the above embodiments have the following beneficial effects: Since the weather forecast data includes not only wind speed, but also the influence of factors such as wave height and water flow velocity, these factors have both instantaneous effects on the ship speed (such as local turbulence) and long-term continuous effects, and the influence of both needs to be considered comprehensively. Therefore, through spatio-temporal feature gating, the temporal correlation of multi-source data in the first weather forecast data is fused to obtain spatio-temporal features, and this data is incorporated into the short-term memory unit, so that the long short-term memory network model can quickly respond to sudden perturbations; further, through environmental perturbation gating, the degree of perturbation of different instantaneous environmental changes on the ship speed is identified from the historical mean of the weather forecast, and the fusion ratio of the feature information of the short-term memory unit and the long-term memory unit is adjusted according to the degree of perturbation, so as to ensure the accurate step of the influence degree of instantaneous changes and continuous changes on the ship speed, and improve the accuracy of the hidden state of the subsequent output to correct the first ship speed.

[0037] Further, determining the first sailing time and the first sailing energy consumption according to the first decision variable, the second decision variable, and the corrected ship speed respectively includes:

[0038] Calculating the second sailing time required for the ship to pass through any two adjacent first preset variable speed points according to the corrected ship speed;

[0039] Determining the first sailing time according to the sum of all the second sailing times; <id=

[0040] Determining the second sailing energy consumption during each of the second sailing times according to the second decision variable;

[0041] Determining the first sailing energy consumption according to the sum of all the second sailing energy consumptions.

[0042] Compared with the prior art, the above embodiments have the following beneficial effects: Since the ship speed needs to be adjusted after passing through each first preset variable speed point, the sailing time between every two adjacent first preset variable speed points needs to be calculated segment by segment, and the sailing energy consumption is calculated according to the sailing time of each segment combined with the power output of each segment, so as to ensure the accuracy of subsequent problem optimization.

[0043] Further, constructing a multi-objective optimization problem according to the first optimization target and the second optimization target includes:

[0044] Taking the first optimization target and the second optimization target as the optimization targets of the multi-objective optimization problem;

[0045] Constructing the constraint conditions of the multi-objective optimization problem according to the power simulation model.

[0046] Further, the mathematical model of the multi-objective optimization problem includes:

[0047]

[0048]

[0049]

[0050] Among them, represents the first sailing time; represents the first sailing energy consumption; represents the randomly generated docking sequence , represents deleting the docking point from the set ; , represents deleting the docking point from the set ; , is the total number of docking points; represents the set of variable-speed points on a certain optimal sailing route, is related to the docking point and the docking point ; represents when at the docking point to the docking point corresponding to the optimal sailing route, the distance from the variable-speed point to the variable-speed point ; represents executing the optimal sailing route from the docking point to the docking point , represents not executing the optimal sailing route from the docking point to the docking point ; represents when at the docking point to the docking point corresponding to the optimal sailing route, the power output of the ship when located at the variable-speed point ; represents the weather forecast data when the ship is at the docking point to the docking point corresponding to the optimal sailing route; is the speed correction operator; is the power simulation model operator; and respectively represent the minimum power output limit and the maximum power output limit; and respectively represent the minimum speed limit and the maximum speed limit.

[0051] Compared with the prior art, the above embodiments have the following beneficial effects: The speed correction operator and the dynamic simulation model operator are respectively embedded in the constructed first optimization objective and the second optimization objective to achieve an accurate mapping between the power output and the speed, so as to achieve the synchronous and precise optimization of the sailing time and the sailing energy consumption; further, the speed correction operator and the dynamic simulation model operator are embedded in the constraint conditions to limit the speed and ensure the effectiveness of the final power output scheme; finally, each docking point is further restricted to dock only once in the constraint conditions to optimize the solution space of the first decision variable, avoid the appearance of invalid solutions at the same time, and improve the optimization efficiency.

[0052] Another embodiment of the present application further provides a ship sailing speed optimization control device based on multi-objective optimization, including: a data acquisition module, a decision variable determination module, a speed correction module, an optimization objective determination module, and a mathematical model solving module;

[0053] Among them, the data acquisition module is used to acquire the sailing route data to be planned and the weather forecast data; the sailing route data includes several docking points, the optimal sailing routes between each docking point, and several preset variable speed points corresponding to each of the optimal sailing routes; the preset variable speed points are set according to the turning points of the corresponding optimal sailing routes.

[0054] The decision variable determination module is used to use the arrival order of each docking point as the first decision variable and the power output of each preset variable speed point corresponding to each optimal sailing route as the second decision variable;

[0055] The speed correction module is used to determine the speeds corresponding to several preset variable speed points according to the first decision variable, the second decision variable, and the preset dynamic simulation model, and correct the speeds by combining the weather forecast data through the preset long short-term memory network model; the dynamic simulation model is a mapping relationship between the power output and the speed built under the condition of no environmental disturbance.

[0056] The optimization objective determination module is used to determine the first sailing time and the first sailing energy consumption respectively according to the first decision variable, the second decision variable, and the corrected speed, and use the first sailing time and the first sailing energy consumption as the first optimization objective and the second optimization objective respectively;

[0057] The mathematical model solving module is used to construct a multi-objective optimization problem according to the first optimization objective and the second optimization objective, and solve the multi-objective optimization problem through a multi-objective optimization algorithm to obtain an optimal power output scheme.

[0058] Further, determining the speeds corresponding to a plurality of preset speed change points according to the first decision variable, the second decision variable, and a preset power simulation model includes:

[0059] Determining the arrival order of each of the docking points according to the first decision variable;

[0060] Determining a plurality of first optimal navigation routes according to the arrival order of each of the docking points;

[0061] Taking the preset speed change points corresponding to the first optimal navigation route as the first preset speed change points;

[0062] Mapping the second decision variable corresponding to each of the first preset speed change points to the corresponding speed through the simulation model. BRIEF DESCRIPTION OF THE DRAWINGS

[0063] In order to more clearly illustrate the technical solutions of the present application, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0064] Figure 1 It is a schematic flow chart of a ship navigation speed optimization control method based on multi-objective optimization provided in some embodiments of the present application;

[0065] Figure 2 It is a schematic structural diagram of a ship navigation speed optimization control device based on multi-objective optimization provided in some embodiments of the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0066] To make the objectives, technical solutions, and advantages of the present application clearer, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are some, but not all, of the embodiments of the present application. All other embodiments obtained by those of ordinary skill in the art without creative efforts based on the embodiments of the present application belong to the scope of protection of the present application.

[0067] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those of ordinary skill in the technical field to which this application belongs; the terms used herein are only for the purpose of describing specific embodiments and are not intended to limit this application; the terms "including" and "having" and any variations thereof in the specification and claims of this application and the above drawings are intended to cover non-exclusive inclusion.

[0068] In the description of the embodiments of the present application, technical terms such as "first" and "second" are only used to distinguish different objects, and cannot be understood as indicating or implying relative importance or implicitly specifying the quantity, specific order, or primary-secondary relationship of the indicated technical features. In the description of the embodiments of the present application, the meaning of "a plurality of" is more than two, unless otherwise specifically defined.

[0069] Reference to "embodiment" herein means that a particular feature, structure, or characteristic described in connection with the embodiment can be included in at least one embodiment of the present application. The phrase appears in various places in the specification and does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment mutually exclusive with other embodiments. Those skilled in the art will explicitly and implicitly understand that the embodiments described herein can be combined with other embodiments.

[0070] In the description of the embodiments of the present application, the term "and / or" is merely a description of the association relationship of associated objects, indicating that three relationships can exist. For example, A and / or B can represent three situations: A exists alone, A and B exist simultaneously, and B exists alone. Additionally, the character " / " in this text generally represents an "or" relationship between the associated objects before and after.

[0071] In the description of the embodiments of the present application, the term "a plurality of" refers to more than two (including two). Similarly, "a plurality of groups" refers to more than two groups (including two groups), and "a plurality of pieces" refers to more than two pieces (including two pieces).

[0072] In the description of the embodiments of the present application, unless otherwise clearly specified and limited, technical terms such as "installation", "connection", "coupling", "fixation", etc. should be understood in a broad sense. For example, it can be a fixed connection, a detachable connection, or integrated; it can also be a mechanical connection or an electrical connection; it can be directly connected or indirectly connected through an intermediate medium, and it can be the communication inside two elements or the interaction relationship between two elements. For those of ordinary skill in the art, the specific meanings of the above terms in the embodiments of the present application can be understood according to specific circumstances.

[0073] In the prior art, generally, a physical process simulation model is constructed, and a weather forecast model is embedded into the physical process simulation model to predict the ship speed under different environments and different power outputs. At the same time, a large number of sensors are set to obtain more accurate real-time weather data to ensure the accuracy of the physical process simulation model. However, the physical process simulation model used in this method has a large number of parameters, requires a large amount of observation data to adjust the parameters of the physical process simulation model, and has a high hardware cost, making it difficult to popularize for different ship types. At the same time, when the physical process simulation model estimates the influence of weather environment factors on the ship speed, it mainly focuses on the instantaneous change influence and fails to consider the influence of persistent weather factors on the ship speed, resulting in higher energy consumption and longer time consumption of the subsequent optimized solutions.

[0074] Please refer to Figure 1 To solve the problem in the prior art of how to more accurately integrate weather forecast data into the ship speed optimization problem while ensuring the computational complexity, thereby reducing the sailing time and sailing energy consumption of the ship speed control scheme, an optimized control method for ship sailing speed based on multi-objective optimization provided by an embodiment of the present application includes S101 to S105, specifically as follows:

[0075] S101: Obtain the sailing route data to be planned and the weather forecast data; the sailing route data includes several stopping points, the optimal sailing routes between each two stopping points, and several preset variable speed points corresponding to each of the optimal sailing routes; the preset variable speed points are set according to the turning points of the corresponding optimal sailing routes.

[0076] Preferably, in some embodiments of the present application, the preset variable speed points are set according to the turning points of the corresponding optimal sailing routes, specifically: read the longitude and latitude coordinates of each optimal sailing route, and further use the coordinate points where the change degree of the front and rear longitude and latitude coordinates exceeds a certain threshold as the variable speed points.

[0077] Preferably, in some embodiments of the present application, when the exact sailing route is already known in advance, in addition to the preset variable speed points being set according to the turning points of the corresponding optimal sailing routes, weather forecast data can be further introduced to predict in advance the points where large weather fluctuations are likely to occur in the sailing route as the preset variable speed points.

[0078] Furthermore, in some embodiments of the present application, the weather forecast data is selected according to a preset interval, and this interval is determined according to the maximum duration allowed for the current sailing.

[0079] S102: Take the arrival order of each of the stopping points as the first decision variable, and take the power output of each of the preset variable speed points corresponding to each of the optimal sailing routes as the second decision variable.

[0080] Further, in some embodiments of the present application, when the exact sailing route has been known in advance, that is, the arrival order of the exact docking points has been known, the first decision variable will no longer be needed at this time, and the second decision variable will become the power output corresponding to each preset variable-speed point under the exact sailing route. That is, the multi-objective optimization problem defined subsequently in the present application can not only solve the optimization problem of the power output scheme for the incomplete planned sailing route, but also optimize the power output scheme based on the already planned sailing path, thereby improving the flexibility of the ship speed optimization control.

[0081] S103: According to the first decision variable, the second decision variable, and a preset power simulation model, determine the ship speeds corresponding to a number of preset variable-speed points, and correct the ship speeds by combining the weather forecast data through a preset long short-term memory network model; the power simulation model is a mapping relationship between power output and ship speed established under the condition of no environmental disturbance.

[0082] Further, in some embodiments of the present application, the determining the ship speeds corresponding to a number of preset variable-speed points according to the first decision variable, the second decision variable, and a preset power simulation model includes:

[0083] Determine the arrival order of each of the docking points according to the first decision variable;

[0084] Determine a number of first optimal sailing routes according to the arrival order of each of the docking points;

[0085] Take the preset variable-speed points corresponding to the first optimal sailing route as the first preset variable-speed points;

[0086] Map the second decision variable corresponding to each of the first preset variable-speed points to the corresponding ship speed through the simulation model.

[0087] To avoid the problem of excessive computational complexity in the optimization process caused by too many power output decisions for all preset variable-speed points in the entire solution space, the optimization space of the subsequent second decision variable is limited by the solution space defined by the first decision variable, thereby effectively improving the optimization speed; further, the second decision variable corresponding to each of the first preset variable-speed points is mapped through the simulation model, so that the sailing speed of the ship at each first preset variable-speed point can be initially determined.

[0088] It can be understood that after the first optimal sailing route is determined, it is equivalent to fixing the first preset variable-speed points passed by the ship in sequence and the passing order of the first preset variable-speed points. At this time, only the power output of these first preset variable-speed points needs to be determined. Therefore, there is no need to perform an optimization operation on all the preset variable-speed points of all the optimal sailing routes. Therefore, in some embodiments of the present application, when the exact sailing route is known in advance, in the above process, only the second decision variables corresponding to each of the first preset variable-speed points need to be directly mapped to the corresponding sailing speeds through the simulation model.

[0089] Preferably, in some embodiments of the present application, the power simulation model is a mapping relationship between power output and sailing speed built under the condition of no environmental disturbance. Specifically, it is a mapping relationship between power output and the first sailing speed built based on Newton's fluid dynamics law and the hull resistance square law, that is, it is equivalent to a simulation model of the ship in a sailing environment without wind and waves.

[0090] Further, in some embodiments of the present application, the method of correcting the sailing speed by combining the preset long short-term memory network model with the weather forecast data includes:

[0091] Obtain the first weather forecast data of each of the first optimal sailing routes within a preset time interval from the weather forecast data;

[0092] Determine the sailing order of each of the first optimal sailing routes according to the first decision variable;

[0093] In the sailing order, combine the first weather forecast data through the preset long short-term memory network model to sequentially correct the sailing speeds corresponding to each of the first preset variable-speed points.

[0094] Since the weather forecast has a time attribute, except for the initial departure time, the time when the ship arrives at each of the first preset variable-speed points subsequently cannot be predicted in advance. Therefore, after determining several first optimal sailing routes, it is necessary to further consider the passing order of each first optimal sailing route. Starting from the first first preset variable-speed point (i.e., the initial starting point), combined with the first weather forecast data at the initial time, gradually iterate to determine the time period of the first weather forecast data required to correct the next first preset variable-speed point, so as to improve the accuracy of sailing speed correction.

[0095] Further, in some embodiments of the present application, since the weather forecast data not only has a time attribute but also a geographical attribute, when obtaining the first weather forecast data, it is not only necessary to ensure that the currently selected time interval can cover all the weather forecast data used for subsequent speed correction, but also the selected data should be the weather forecast data for the area corresponding to the current first optimal navigation route. Therefore, the actual first weather forecast data is determined by continuous updating during the subsequent iterative correction process. It can be understood that the first weather forecast data is a function related to time variables and geographical variables, and the time variables and geographical variables are related to the speed. By gradually correcting the speed, the accuracy of the first weather forecast data used subsequently is further gradually corrected to ensure the accuracy of the entire correction process.

[0096] Further, in some embodiments of the present application, the step of sequentially correcting the speeds corresponding to each first preset speed change point according to the navigation order by combining the first weather forecast data with a preset long short-term memory network model includes:

[0097] Taking the speed to be corrected currently as the first speed, and taking the time point when the ship reaches the first preset speed change point corresponding to the first speed as the first time point, and judging whether the first preset speed change point corresponding to the first speed is the first preset speed change point corresponding to any of the first optimal navigation routes;

[0098] If so, correct the first speed by combining the first weather forecast data at the first time point through a data-driven method;

[0099] Otherwise, taking the time point when the ship reaches the first preset speed change point corresponding to the speed corrected last time as the second time point, and correcting the first speed through the preset long short-term memory network model according to the first weather forecast data during the period from the first time point to the second time point.

[0100] Since the ship starts from each docking point at the first first preset speed change point of each first optimal navigation route, and there are no accumulated weather environment factors affecting before this first preset speed change point, it is necessary to correct the initial first speed with the first weather forecast data in the form of a time point; when the ship leaves the first first preset speed change point, it starts to be affected by the accumulated weather environment factors. Therefore, according to the first weather forecast data (at this time, the data is time series data within a time period) experienced between the first preset speed change points of two vectors, the feature information in the time series data is obtained through the preset long short-term memory network model to correct the first speed, thereby improving the correction accuracy.

[0101] Preferably, in some embodiments of the present application, by means of a data-driven method, in combination with the first weather forecast data at the first time point, the first sailing speed is corrected, including: wind speed correction, wave height correction, and ocean current correction;

[0102] Among them, the formula for wind speed correction is:

[0103]

[0104] The formula for wave height correction is:

[0105]

[0106] The formula for ocean current correction is:

[0107]

[0108] Then the finally corrected first sailing speed is:

[0109]

[0110] Among them, 、 and are the correction amounts of wind speed, wave height, and ocean current to the first sailing speed respectively; and are the correction coefficients of wind speed and wave height respectively; is the wind speed; is the wind direction; is the ship form course; is the wave height; is the wave frequency; is the time; is the ocean current speed; is the ocean current direction.

[0111] Furthermore, in some embodiments of the present application, the hidden state recurrence process of the preset long short-term memory network model includes:

[0112] Among them, the preset long short-term memory network model includes: spatio-temporal feature gating, environmental disturbance gating, short-term memory unit, and long-term memory unit;

[0113] Through the spatio-temporal feature gating, spatio-temporal features are extracted from the first weather forecast data at the current time step, and the short-term memory unit is updated through the spatio-temporal features;

[0114] Fusing the feature information of the long-term memory unit at the previous time step and the feature information of the short-term memory unit at the current time step to obtain the feature information of the long-term memory unit at the current time step;

[0115] Through the environmental disturbance gating, combined with the historical average of weather forecasts, obtain the disturbance weight;

[0116] According to the disturbance weight and the feature information of the short-term memory unit and the long-term memory unit at the current time step, update the memory cells of the preset long short-term memory network model;

[0117] According to the updated memory cells, obtain the hidden state output by the preset long short-term memory network model at the current time step.

[0118] Specifically, in some embodiments of the present application, the recurrence formula for the hidden state of the preset long short-term memory network model includes:

[0119]

[0120]

[0121]

[0122]

[0123]

[0124]

[0125]

[0126]

[0127]

[0128]

[0129]

[0130]

[0131]

[0132] Among them, is the forgetting gate vector at the current time step; is the Sigmoid function; 、 、 、 、 、 and are the weight matrices for each update step; 、 、 、 , , and are the bias terms for each update step; is the input data for the current time step; and are the final output vectors for the previous time step and the current time step respectively; is the input gate vector for the current time step; is the spatio-temporal feature gate vector for the current time step; is the environmental disturbance gate; is the candidate memory unit for the current time step; and are the short-term memory units for the previous time step and the current time step respectively; and are the long-term memory units for the previous time step and the current time step respectively; is the long-short term memory fusion coefficient; is the memory unit for the current time step; is the short-long memory dynamic fusion weight; is the Softmax function; is the output gate vector for the current time step; is the hyperbolic tangent function; is the exponential decay function; is the element-wise multiplication; is the embedded feature extraction function; are the historical average data of wind speed, wave height and ocean current velocity before the current time step respectively; is the embedded feature for the current time step.

[0133] Furthermore, in some embodiments of the present application, the input data for each time step is:

[0134]

[0135] wherein, represents the wind speed data in the first weather forecast data at the current time step, and this data is vector data (i.e., it needs to include the direction of the wind); is the maximum wind speed; represents the wave height data in the first weather forecast data at the current time step, and this data is also vector data (i.e., it needs to include the direction of the wave, generally the same as the wind direction, but may change in special cases); is the maximum wave height; represents the ocean current velocity data in the first weather forecast data at the current time step, and this data is also vector data (i.e., it needs to include the flow direction of the ocean current); is the maximum ocean current velocity. It should be noted that , and should be the average value of the historical maximum statistical data in each current quarter on the first optimal navigation route where the current ship is located, to prevent the characteristics of the subsequent input data from changing insignificantly due to an overly large maximum value.

[0136] Furthermore, in some embodiments of the present application, when the hidden state of the long short-term memory network model is recursively obtained through the recurrence formula and the hidden state at the final time step is obtained through repeated iteration after that, it is further necessary to correct the first ship speed according to this hidden state , and the specific formula is:

[0137]

[0138]

[0139] where is the correction coefficient; is the first ship speed, and this data is vector data, which is mainly used to introduce the navigation direction of the current ship in the correction coefficient calculation formula; is the offset term; is the corrected first ship speed.

[0140] It can be seen from the above embodiments that since the weather forecast data includes not only wind speed, but also factors such as wave height and water flow velocity, these factors have not only an instantaneous impact on the ship speed (such as local turbulence), but also a long-term continuous impact, and it is necessary to comprehensively consider the impacts of both. Therefore, on the basis of the traditional long short-term memory network model, the present application additionally introduces spatio-temporal feature gating, environmental disturbance gating, short-term memory units, and long-term memory units. In the improved long short-term memory network model, through spatio-temporal feature gating, the time correlation of multi-source data in the first weather forecast data is fused to obtain spatio-temporal features, and this data is incorporated into the short-term memory unit, so that the long short-term memory network model can quickly respond to sudden disturbances; further through environmental disturbance gating, the degree of disturbance of different instantaneous environmental changes on the ship speed is identified from the historical mean of the weather forecast, and the fusion ratio of the feature information of the short-term memory unit and the long-term memory unit is adjusted according to this degree of disturbance, so as to ensure the accurate steps of the impact degrees of instantaneous changes and continuous changes on the ship speed, and improve the accuracy of correcting the first ship speed by the subsequent output hidden state.

[0141] S104: According to the first decision variable, the second decision variable, and the corrected ship speed, respectively determine the first navigation time and the first navigation energy consumption, and use the first navigation time and the first navigation energy consumption as the first optimization target and the second optimization target respectively.

[0142] Further, in some embodiments of the present application, determining the first sailing time and the first sailing energy consumption according to the first decision variable, the second decision variable, and the corrected sailing speed respectively includes:

[0143] Calculating the second sailing time required for the ship to pass through any two adjacent first preset variable speed points according to the corrected sailing speed;

[0144] Determining the first sailing time according to the sum of all the second sailing times;

[0145] Determining the second sailing energy consumption during each of the second sailing times according to the second decision variable;

[0146] Determining the first sailing energy consumption according to the sum of all the second sailing energy consumptions.

[0147] Since the ship needs to adjust its speed after passing through each first preset variable speed point, the sailing time between every two adjacent first preset variable speed points needs to be calculated segment by segment, and the sailing energy consumption is calculated based on the power output of each segment in combination with the sailing time of each segment, so as to ensure the accuracy of subsequent problem optimization.

[0148] It should be noted that the above process of determining the first sailing time and the first sailing energy consumption refers to establishing the functional relationship between the first decision variable and the second decision variable, and the first sailing time and the first sailing energy consumption, rather than directly calculating the actual first sailing time and the actual first sailing energy consumption.

[0149] S105: Construct a multi-objective optimization problem according to the first optimization target and the second optimization target, and solve the multi-objective optimization problem through a multi-objective optimization algorithm to obtain an optimal power output scheme.

[0150] Further, in some embodiments of the present application, constructing a multi-objective optimization problem according to the first optimization target and the second optimization target includes:

[0151] Taking the first optimization target and the second optimization target as the optimization targets of the multi-objective optimization problem;

[0152] Constructing the constraint conditions of the multi-objective optimization problem according to the power simulation model.

[0153] Further, in some embodiments of the present application, the mathematical model of the multi-objective optimization problem includes:

[0154]

[0155]

[0156]

[0157] Among them, represents the first voyage time; represents the first voyage energy consumption; represents the randomly generated docking sequence , represents deleting the docking point from the set , represents deleting the docking point from the set , is the total number of docking points; represents the set of variable-speed points on a certain optimal voyage route, is related to the docking point and the docking point ; represents when on the optimal voyage route from the docking point to the docking point , the distance from the variable-speed point to the variable-speed point ; represents when executing the optimal voyage route from the docking point to the docking point , represents not executing the optimal voyage route from the docking point to the docking point ; represents the power output of the ship when it is at the variable-speed point to the docking point on the corresponding optimal voyage route; represents the weather forecast data of the ship when it is at the docking point from the docking point to the docking point on the corresponding optimal voyage route; is the speed correction operator; is the power simulation model operator; and respectively represent the minimum power output limit and the maximum power output limit; and respectively represent the minimum speed limit and the maximum speed limit.

[0158] ​​Embed the speed correction operator and the dynamic simulation model operator into the first optimization objective and the second optimization objective respectively to achieve an accurate mapping between power output and speed, so as to synchronously and accurately optimize the sailing time and sailing energy consumption; further embed the speed correction operator and the dynamic simulation model operator into the constraint conditions to limit the speed and ensure the effectiveness of the final power output scheme; finally, further limit each docking point to only dock once in the constraint conditions to optimize the solution space of the first decision variable, avoid the appearance of invalid solutions, and improve the optimization efficiency.

[0159] Further, in some embodiments of the present application, solving the multi-objective optimization problem through a multi-objective optimization algorithm to obtain an optimal power output scheme includes, but is not limited to, multi-objective optimization algorithms such as genetic algorithms, particle swarm optimization algorithms, and simulated annealing algorithms.

[0160] Preferably, in some embodiments of the present application, in order to improve the optimization efficiency and optimization effect, the mutation operator and the crossover operator in the genetic algorithm are next improved. Specifically, at the beginning of iteration, increase the mutation probability of the first several preset genes of each chromosome, and only perform crossover operations on the first several preset genes; further, when the iteration result tends to be stable, fix the first several preset genes and only perform crossover and mutation operations on the remaining genes.

[0161] Since in the ship speed optimization problem, weather forecast data is the core key influencing factor, and it is also known that the accuracy of the current speed correction is related to the accuracy of the current weather forecast data selection, and the accuracy of the current weather forecast data selection is related to the accuracy of the previously corrected speed, so in fact it represents that on a sailing route, the power output decision corresponding to the preset speed change point passed earlier will affect the power output decision corresponding to the preset speed change point passed later to a greater extent.

[0162] Based on the above analysis results, it can be known that by focusing on searching for the power output decision of the preset speed change point passed earlier in the early stage of iteration, a large number of invalid searches can be avoided. For example, for a complete power output decision scheme, the output decision corresponding to the preset speed change point passed later reaches an optimal state, but if the output decision corresponding to the preset speed change point passed earlier in the power output decision scheme is changed during the iteration process, then the power output decision scheme obtained after a large number of iterations is likely to be damaged in the optimal state, thus affecting the optimization efficiency.

[0163] In summary, it can be seen that a ship navigation speed optimization control method based on multi-objective optimization provided by an embodiment of the present application has the following beneficial effects: Whenever the ship sails to the turning point of the optimal navigation route, it is easier for the influence of the weather environment on the sailing speed to change. Therefore, by setting the turning point of each optimal navigation route as a preset speed change point, the sensitivity of subsequent speed optimization to weather forecast data is improved; further, through a simplified dynamic simulation model, while solving the problem of low versatility caused by the overly complex physical process model, a long short-term memory network model that can capture the change characteristics of time series data is introduced to further correct the sailing speed. The above process of estimating the sailing speed through the weather environment not only reduces the computational complexity but also ensures the accuracy of the sailing speed estimation; finally, by establishing a multi-objective optimization problem and solving the optimal power output scheme, the final power output scheme can reduce both the sailing time and the sailing energy consumption during the sailing process.

[0164] As Figure 2 shown, based on the above method item embodiment, an embodiment of the present application provides a ship navigation speed optimization control device based on multi-objective optimization, including: a data acquisition module 201, a decision variable determination module 202, a sailing speed correction module 203, an optimization target determination module 204, and a mathematical model solving module 205.

[0165] Further, in some embodiments of the present application, the data acquisition module 201 is configured to acquire the navigation route data to be planned and the weather forecast data; the navigation route data includes a number of stop points, the optimal navigation routes between the stop points, and a number of preset speed change points corresponding to each of the optimal navigation routes; the preset speed change points are set according to the turning points of the corresponding optimal navigation routes; the decision variable determination module 202 is configured to use the arrival order of each of the stop points as the first decision variable and the power output of each of the preset speed change points corresponding to each of the optimal navigation routes as the second decision variable; the sailing speed correction module 203 is configured to determine the sailing speeds corresponding to a number of preset speed change points according to the first decision variable, the second decision variable, and a preset dynamic simulation model, and correct the sailing speed by combining the weather forecast data with a preset long short-term memory network model; the dynamic simulation model is a mapping relationship between power output and sailing speed built under the condition of no environmental disturbance; the optimization target determination module 204 is configured to determine the first sailing time and the first sailing energy consumption according to the first decision variable, the second decision variable, and the corrected sailing speed, and use the first sailing time and the first sailing energy consumption as the first optimization target and the second optimization target respectively; the mathematical model solving module 205 is configured to construct a multi-objective optimization problem according to the first optimization target and the second optimization target, and solve the multi-objective optimization problem through a multi-objective optimization algorithm to obtain an optimal power output scheme.

[0166] Further, in some embodiments of the present application, determining the speeds corresponding to several preset speed change points according to the first decision variable, the second decision variable, and a preset power simulation model includes: determining the arrival order of each of the docking points according to the first decision variable; determining several first optimal navigation routes according to the arrival order of each of the docking points; taking the preset speed change points corresponding to the first optimal navigation routes as the first preset speed change points; and mapping the second decision variable corresponding to each of the first preset speed change points to the corresponding speed through the simulation model.

[0167] Further, in some embodiments of the present application, correcting the speed by combining the preset long short-term memory network model with the weather forecast data includes: obtaining the first weather forecast data of each of the first optimal navigation routes within a preset time interval from the weather forecast data; determining the navigation order of each of the first optimal navigation routes according to the first decision variable; and sequentially correcting the speeds corresponding to each of the first preset speed change points by combining the first weather forecast data through the preset long short-term memory network model in the navigation order.

[0168] Further, in some embodiments of the present application, sequentially correcting the speeds corresponding to each of the first preset speed change points by combining the first weather forecast data through the preset long short-term memory network model in the navigation order includes: taking the speed to be corrected currently as the first speed, taking the time point when the ship arrives at the first preset speed change point corresponding to the first speed as the first time point, and determining whether the first preset speed change point corresponding to the first speed is the first preset speed change point corresponding to any of the first optimal navigation routes; if so, correcting the first speed by a data-driven method in combination with the first weather forecast data at the first time point; otherwise, taking the time point when the ship arrives at the first preset speed change point corresponding to the previously corrected speed as the second time point, and correcting the first speed through the preset long short-term memory network model according to the first weather forecast data during the period from the first time point to the second time point.

[0169] Further, in some embodiments of the present application, the hidden state recurrence process of the preset long short-term memory network model includes: wherein, the preset long short-term memory network model includes: a spatio-temporal feature gating, an environmental disturbance gating, a short-term memory unit, and a long-term memory unit; through the spatio-temporal feature gating, spatio-temporal features are extracted from the first weather forecast data at the current time step, and the short-term memory unit is updated through the spatio-temporal features; the feature information of the long-term memory unit at the previous time step and the feature information of the short-term memory unit at the current time step are fused to obtain the feature information of the long-term memory unit at the current time step; through the environmental disturbance gating, the disturbance weight is obtained in combination with the historical average of the weather forecast; according to the disturbance weight and the feature information of the short-term memory unit and the long-term memory unit at the current time step, the memory cell of the preset long short-term memory network model is updated; according to the updated memory cell, the hidden state output by the preset long short-term memory network model at the current time step is obtained.

[0170] Further, in some embodiments of the present application, the determining the first sailing time and the first sailing energy consumption according to the first decision variable, the second decision variable, and the corrected sailing speed includes: calculating the second sailing time required for the ship to pass through any two adjacent first preset variable speed points according to the corrected sailing speed; determining the first sailing time according to the sum of all the second sailing times; determining the second sailing energy consumption during each of the second sailing times according to the second decision variable; determining the first sailing energy consumption according to the sum of all the second sailing energy consumptions.

[0171] Further, in some embodiments of the present application, the constructing a multi-objective optimization problem according to the first optimization target and the second optimization target includes: taking the first optimization target and the second optimization target as the optimization targets of the multi-objective optimization problem; constructing the constraint conditions of the multi-objective optimization problem according to the dynamic simulation model.

[0172] Further, in some embodiments of the present application, the mathematical model of the multi-objective optimization problem includes:

[0173]

[0174]

[0175]

[0176] Wherein, represents the first sailing time; represents the first sailing energy consumption; represents the randomly generated docking sequence , Represents deleting a stop point from the set , , Represents deleting a stop point from the set , , where is the total number of stop points; Represents the set of variable-speed points on a certain optimal navigation route, which is related to the stop point and the stop point ; Represents the distance from the variable-speed point to the variable-speed point when corresponding to the optimal navigation route between the stop point and the stop point ; Represents executing the optimal navigation route from the stop point to the stop point , represents not executing the optimal navigation route from the stop point to the stop point ; Represents the power output of the ship when it is at the variable-speed point when corresponding to the optimal navigation route between the stop point and the stop point ; Represents the weather forecast data when the ship is at the stop point to the stop point corresponding to the optimal navigation route; is the speed correction operator; is the power simulation model operator; and respectively represent the minimum power output limit and the maximum power output limit; and respectively represent the minimum speed limit and the maximum speed limit.

[0177] It can be understood that the above device item embodiments correspond to the method item embodiments of the present application, and can implement the method for optimizing the ship navigation speed based on multi-objective optimization provided by any one of the above method item embodiments of the present application.

[0178] It should be noted that the above-described device embodiments are only illustrative, and some or all of the modules can be selected according to actual needs to achieve the purpose of the present embodiment solution. In addition, in the accompanying drawings of the device embodiments provided by the present application, the connection relationship between the modules indicates that they have a communication connection, which can be specifically implemented as one or more communication buses or signal lines. Those of ordinary skill in the art can understand and implement it without creative efforts.

[0179] In summary, it can be seen that a ship navigation speed optimization control device provided by an embodiment of the present application based on multi-objective optimization has the following beneficial effects: Whenever the ship sails to the turning point of the optimal navigation route, it is more likely to have a change in the influence of weather conditions on the sailing speed. Therefore, by setting the turning points of each optimal navigation route as preset variable speed points, the sensitivity of subsequent speed optimization to weather forecast data is improved; further, through a simplified dynamic simulation model, while solving the problem of low versatility caused by the overly complex physical process model, a long short-term memory network model that can capture the change characteristics of time series data is introduced to further correct the sailing speed. The above process of estimating the sailing speed through weather conditions not only reduces the computational complexity but also ensures the accuracy of the sailing speed estimation; finally, by establishing a multi-objective optimization problem and solving the optimal power output scheme, the final power output scheme can reduce both the sailing time and the sailing energy consumption during the sailing process.

[0180] Based on the above embodiment of the ship navigation speed optimization control method based on multi-objective optimization, another embodiment of the present application provides a terminal device, which includes a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, it implements the ship navigation speed optimization control method based on multi-objective optimization according to any embodiment of the present application.

[0181] Exemplarily, in this embodiment, the computer program can be divided into one or more modules, and the one or more modules are stored in the memory and executed by the processor to complete the present application. The one or more module elements can be a series of computer program instruction segments capable of completing specific functions, and the instruction segments are used to describe the execution process of the computer program in the terminal device.

[0182] The terminal device can be a computing device such as a desktop computer, a notebook, a palm computer, and a cloud server. The terminal device may include, but is not limited to, a processor and a memory.

[0183] The so-called processor may be a Central Processing Unit (CPU), or may also be other general-purpose processors, Digital Signal Processors (DSPs), Application Specific Integrated Circuits (ASICs), Field-Programmable Gate Arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc. The processor is the control center of the terminal device and connects various parts of the entire terminal device through various interfaces and circuits.

[0184] Based on the above method item embodiments, another embodiment of the present application provides a computer-readable storage medium, including a stored computer program, wherein when the computer program runs, it controls the device where the computer-readable storage medium is located to execute the method for optimizing the ship navigation speed based on multi-objective optimization described in any one of the above method item embodiments of the present application.

[0185] Among them, if the modules / units integrated in the device / terminal device are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on such an understanding, to implement all or part of the processes in the above method embodiments of the present application, it can also be completed by instructing relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium. When the computer program is executed by a processor, the steps of the above various method embodiments can be implemented. Among them, the computer program includes computer program code, and the computer program code can be in the form of source code, object code, executable file, or some intermediate form, etc. The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disc, computer memory, Read-Only Memory (ROM), Random Access Memory (RAM), electrical carrier signal, telecommunication signal, and software distribution medium, etc.

Claims

1. A ship navigation speed optimization control method based on multi-objective optimization, characterized in that Including: Obtaining the navigation route data to be planned and the weather forecast data; The navigation route data includes several stopping points, the optimal navigation routes between each pair of stopping points, and several preset variable-speed points corresponding to each of the optimal navigation routes; the preset variable-speed points are set according to the turning points of the corresponding optimal navigation routes; Taking the arrival order of each of the stopping points as the first decision variable, and taking the power output of each of the preset variable-speed points corresponding to each of the optimal navigation routes as the second decision variable; According to the first decision variable, the second decision variable, and a preset power simulation model, determining the ship speeds corresponding to several preset variable-speed points, and correcting the ship speeds by combining the weather forecast data through a preset long short-term memory network model; the power simulation model is a mapping relationship between power output and ship speed built under the condition of no environmental disturbance; According to the first decision variable, the second decision variable, and the corrected ship speeds, respectively determining the first navigation time and the first navigation energy consumption, and taking the first navigation time and the first navigation energy consumption as the first optimization target and the second optimization target respectively; According to the first optimization target and the second optimization target, constructing a multi-objective optimization problem, and solving the multi-objective optimization problem through a multi-objective optimization algorithm to obtain an optimal power output plan.

2. The method for optimizing and controlling the ship navigation speed based on multi-objective optimization according to claim 1, wherein, The determining the ship speeds corresponding to several preset variable-speed points according to the first decision variable, the second decision variable, and a preset power simulation model includes: According to the first decision variable, determining the arrival order of each of the stopping points; According to the arrival order of each of the stopping points, determining several first optimal navigation routes; Taking the preset variable-speed points corresponding to the first optimal navigation routes as the first preset variable-speed points; Through the simulation model, mapping the second decision variables corresponding to each of the first preset variable-speed points to the corresponding ship speeds.

3. The method for optimizing the ship navigation speed based on multi-objective optimization according to claim 2, wherein The correcting the ship speeds by combining the weather forecast data through a preset long short-term memory network model includes: Obtaining the first weather forecast data of each of the first optimal navigation routes within a preset time interval from the weather forecast data; According to the first decision variable, determining the navigation order of each of the first optimal navigation routes; In the navigation order, successively correcting the ship speeds corresponding to each of the first preset variable-speed points by combining the first weather forecast data through a preset long short-term memory network model.

4. The optimal control method for ship navigation speed based on multi-objective optimization according to claim 3, wherein, The successively correcting the ship speeds corresponding to each of the first preset variable-speed points by combining the first weather forecast data through a preset long short-term memory network model in the navigation order includes: Taking the ship speed to be currently corrected as the first ship speed, and taking the time point when the ship arrives at the first preset variable-speed point corresponding to the first ship speed as the first time point, and judging whether the first preset variable-speed point corresponding to the first ship speed is the first preset variable-speed point corresponding to any of the first optimal navigation routes; If so, correcting the first ship speed by a data-driven method in combination with the first weather forecast data at the first time point; Otherwise, take the time point when the ship reaches the first preset speed change point corresponding to the previous corrected speed as the second time point, and correct the first speed according to the first weather forecast data during the period from the first time point to the second time point through the preset long short-term memory network model.

5. The method for optimizing the ship navigation speed based on multi-objective optimization according to claim 4, characterized in that, The hidden state recursion process of the preset long short-term memory network model includes: wherein, the preset long short-term memory network model includes: spatio-temporal feature gating, environmental disturbance gating, short-term memory unit and long-term memory unit; Extract spatio-temporal features from the first weather forecast data at the current time step through the spatio-temporal feature gating, and update the short-term memory unit through the spatio-temporal features; Fuse the feature information of the long-term memory unit at the previous time step and the feature information of the short-term memory unit at the current time step to obtain the feature information of the long-term memory unit at the current time step; Obtain the disturbance weight by combining the historical mean of the weather forecast through the environmental disturbance gating; Update the memory cell of the preset long short-term memory network model according to the disturbance weight and the feature information of the short-term memory unit and the long-term memory unit at the current time step; Obtain the hidden state output by the preset long short-term memory network model at the current time step according to the updated memory cell.

6. The method for optimizing and controlling the ship navigation speed based on multi-objective optimization according to claim 3, characterized in that, The determining the first sailing time and the first sailing energy consumption according to the first decision variable, the second decision variable and the corrected speed respectively includes: Calculate the second sailing time required for the ship to pass through any two adjacent first preset speed change points according to the corrected speed; Determine the first sailing time according to the sum of all the second sailing times; Determine the second sailing energy consumption during each of the second sailing times according to the second decision variable; Determine the first sailing energy consumption according to the sum of all the second sailing energy consumptions.

7. The method for optimizing and controlling the ship navigation speed based on multi-objective optimization according to claim 1, characterized in that The constructing a multi-objective optimization problem according to the first optimization target and the second optimization target includes: Take the first optimization target and the second optimization target as the optimization targets of the multi-objective optimization problem; Construct the constraint conditions of the multi-objective optimization problem according to the power simulation model.

8. The optimization control method for ship navigation speed based on multi-objective optimization according to claim 7, wherein, The mathematical model of the multi-objective optimization problem includes: ; ; ; Among them, represents the first sailing time; represents the first sailing energy consumption; represents the randomly generated docking sequence , represents deleting the docking point from the set , represents deleting the docking point from the set ,[[ID=2,1]] is the total number of docking points; represents the set of variable speed points on a certain optimal sailing route, is related to the docking point and the docking point ; represents the distance from the variable speed point to the variable speed point on the optimal sailing route corresponding to the docking point to ; represents when executing the optimal sailing route from the docking point to the docking point , represents not executing the optimal sailing route from the docking point to the docking point ; represents the power output of the ship when it is at the variable speed point to the docking point on the optimal sailing route corresponding to the docking point ; represents the weather forecast data of the ship when it is at the docking point to the docking point on the optimal sailing route corresponding to the docking point is the speed correction operator; is the power simulation model operator; and respectively represent the minimum power output limit and the maximum power output limit; and respectively represent the minimum speed limit and the maximum speed limit.

9. An optimized control device for ship navigation speed based on multi-objective optimization, characterized in that, includes: data acquisition module, decision variable determination module, speed correction module, optimization target determination module and mathematical model solving module; wherein, the data acquisition module is used to acquire the sailing route data to be planned and the weather forecast data; the sailing route data includes several stop points, the optimal sailing routes between each stop point and several preset speed change points corresponding to each of the optimal sailing routes; the preset speed change points are set according to the turning points of the corresponding optimal sailing routes; The decision variable determination module is used to take the arrival order of each stop point as the first decision variable and the power output of each preset speed change point corresponding to each optimal sailing route as the second decision variable; The speed correction module is used to determine the speeds corresponding to a number of preset speed change points according to the first decision variable, the second decision variable, and a preset power simulation model, and correct the speeds by combining the weather forecast data with a preset long short-term memory network model; the power simulation model is a mapping relationship between power output and speed built under the condition of no environmental disturbance; The optimization objective determination module is used to respectively determine the first sailing time and the first sailing energy consumption according to the first decision variable, the second decision variable, and the corrected speed, and use the first sailing time and the first sailing energy consumption as the first optimization objective and the second optimization objective respectively; The mathematical model solving module is used to construct a multi-objective optimization problem according to the first optimization objective and the second optimization objective, and solve the multi-objective optimization problem through a multi-objective optimization algorithm to obtain an optimal power output scheme.

10. The optimized control device for the ship navigation speed based on multi-objective optimization according to claim 9, characterized in that, Determining the speeds corresponding to a number of preset speed change points according to the first decision variable, the second decision variable, and a preset power simulation model includes: Determining the arrival order of each of the docking points according to the first decision variable; Determining a number of first optimal sailing routes according to the arrival order of each of the docking points; Taking the preset speed change points corresponding to the first optimal sailing route as the first preset speed change points; Mapping the second decision variable corresponding to each of the first preset speed change points to the corresponding speed through the simulation model.

Images