CN112550592A - Data-driven ship energy consumption prediction method - Google Patents

Abstract

The invention belongs to the technical field of ship energy consumption prediction, and particularly relates to a ship energy consumption prediction method based on data driving. The method comprises the steps of obtaining expected course, departure time and sea condition information of a ship; constructing an off-line basic model, which comprises a ship water steady state navigational speed model, an engine power model and an engine oil consumption model; carrying out high-dimensional linear interpolation on discrete grid matrixes of the ship to water steady-state navigational speed model and the engine power model; performing one-dimensional linear interpolation on a discrete grid matrix of the engine oil consumption model; and deducing a ship energy consumption model in real time. The invention takes a set of data matrix as support to realize the rapid calculation of the energy consumption of the long navigation path of the ship. The method does not influence the idea of the original ship dynamics simulation model, coexists with the dynamics simulation mode in the actual simulation verification and evaluation system, and is respectively responsible for different types of simulation calculation.

Description

Data-driven ship energy consumption prediction method

Technical Field

The invention belongs to the technical field of ship energy consumption prediction, and particularly relates to a ship energy consumption prediction method based on data driving.

Background

The rapid development of national economy in China has the advantages that globalization pace in various fields is faster and faster, ship transportation construction is more and more important to the aspects of society, economy and the like in China, ship energy consumption is an important factor of ship transportation economy, the development of ship transportation is directly related, and meanwhile, the influence of the ship energy consumption on the aspects of ship management efficiency, environmental pollution and the like is large. Therefore, the research and analysis on the ship energy consumption have important theoretical research and engineering practice significance, and particularly in the aspect of ship energy consumption prediction, the effective ship energy consumption prediction has wider industrial prospect and commercial space.

Scholars and engineers at home and abroad adopt different methods and propose ship energy consumption prediction methods from different angles, for example, Liuyi Fan in the treatise on modeling and simulation analysis of ship energy efficiency operation index prediction establishes a complex mathematical model of a propulsion system, ship motion and environment, and iterative operation is carried out on the basis to solve oil consumption. For example, in the treatise on the research on the ship energy consumption assessment method and energy consumption prediction in Lin Rong Mode, each unit of ship energy consumption is established, and various mathematical method theories are adopted to deduce an energy consumption value. The above method has several problems: firstly, a ship dynamics model is excessively relied on, but an accurate mathematical model is difficult to establish due to the fact that the ship power energy conversion process has multiple links and complex influence factors, and model parameters have great influence on the ship energy consumption prediction effect, so that ship energy consumption evaluation and prediction cannot be effectively carried out; secondly, the mathematical models of different ships are different, and the method parameter settings in the energy consumption prediction process are also different, so that the universality of the energy consumption prediction method based on the ship dynamics model is poor; thirdly, when a long flight path (such as a flight path with a flight time of several days) can be predicted, the calculation amount is huge, and the period of the dynamic model in combination with the test result to correct the model is long, so that the actual ship application requirements are sometimes difficult to meet.

Disclosure of Invention

The invention aims to provide a ship energy consumption prediction method based on data driving, which can predict ship energy consumption by using a data driving method under the condition of not accurately using a ship energy consumption mathematical model.

The purpose of the invention is realized by the following technical scheme: the method comprises the following steps:

step 1: acquiring expected route and departure time T of ship₀And sea state information;

step 2: constructing an off-line basic model, which comprises a ship water steady state navigational speed model, an engine power model and an engine oil consumption model;

the ship water-alignment steady state navigational speed model is composed of a 4-dimensional matrix V_w(φ_w,H_w,T_w,N_D) Represents; v_wRepresenting the steady state navigational speed of the ship to water; phi is a_wRepresenting the wave angle relative to the vessel's own system; h_wRepresents a wave height value; t is_wRepresents the wave period value; n is a radical of_DIs the desired speed of the engine; the engine power model is composed of a 4-dimensional matrix P_m(φ_w,H_w,T_w,V_w) Represents; p_mRepresenting engine power; the engine oil consumption model is composed of a 1-dimensional matrix O_m(P_m) Represents; o is_mRepresenting engine fuel consumption rate;

the three models need to acquire data through a ship pool test or a real ship test, relationship data between respective variables and function values are represented in a normalized grid form, each dimension of matrix input data is sampled into a plurality of points at certain intervals, but a plurality of dimensions are orthogonal to each other to form normalized grid data;

and step 3: carrying out high-dimensional linear interpolation on discrete grid matrixes of the ship to water steady-state navigational speed model and the engine power model; performing one-dimensional linear interpolation on a discrete grid matrix of the engine oil consumption model;

the ship water steady-state navigational speed model and the engine power model are in a 4-dimensional form, and the high-dimensional linear interpolation principle is consistent;

the method for performing high-dimensional linear interpolation on the water steady-state navigational speed model of the ship specifically comprises the following steps:

step 3.1: selection matrix V_w＝V_w(φ_w,H_w,T_w,N_D) Medium desired speed N_DFor data of the same value, for wave height value H_wWave angle phi relative to the ship's own system_wWave period value T_wCarrying out three-dimensional linear interpolation;

step 3.2: the ship water steady state navigational speed model obtained after the three-dimensional linear interpolation and the expected rotating speed N of the engine_DCarrying out two-dimensional linear interpolation to obtain a high-dimensional linear interpolation ship water-steady state navigational speed linear model, and calculating a ship water-steady state navigational speed value under the condition that the wave direction angle, the wave height and the wave period of any ship body system and the expected rotating speed of an engine are known;

and 4, step 4: deducing a ship energy consumption model in real time;

step 4.1: interpolating the expected ship route to form a high-density path point row, ensuring that the maximum distance of each path point is within a set threshold value, wherein the generated path point row contains N path points and N-1 path subsections; calculating longitude and latitude coordinates of each interpolation point in a one-dimensional linear interpolation mode between two route points according to the known route point information, and distributing the known expected engine speed information for each interpolation route point;

step 4.2: traversing each path subsection, and calculating the oil consumption of each path subsection;

step 4.2.1: according to the K-1 th path sub-section L_K-1Time of arrival t_LK-1And L_K-1Acquiring the K path subsection L according to the longitude and latitude information of the end point position of the section_KSea state information of (1), including wave height H_wWave direction and wave period T_w，K＝1,2，...，N-1；

Step 4.2.2: according to the Kth path sub-section L_KCalculating a two-point connecting line vector by utilizing the longitude and latitude information of the starting point and the end point, and calculating the path length and the direction angle of the path vector, wherein the angle is used as the direction angle of the ground speed of the ship;

step 4.2.3: according to L_KDesired engine speed N of the segment_LK-1And the acquired wave height, wave direction and wave period data are substituted into a water steady state navigational speed model V of the ship_w＝V_w(φ_w,H_w,T_w,N_D) Calculating the water speed V of the ship_w；

It is necessary to mix L_KThe wave direction of the segment is transformed into a wave direction angle phi relative to the ship's own train_wReplacing the direction angle of the ship to the water speed with the direction angle of the ship to the ground speed during conversion;

step 4.2.4: will ship to water speed V_wSubstituting sea state information into engine power model P_m＝P_m(φ_w,H_w,T_w,V_w) Calculate L_KThe power of the section engine is substituted into the engine oil consumption model O_m＝O_m(P_m) Calculating the oil consumption speed of the ship engine;

step 4.2.5: according to known L_KSpeed and direction angle L of section-to-ground navigation_KSegment pair water velocity, flow velocity and flow velocity direction angle, and calculating L by using trigonometric relation sine theorem_KSpeed V of section to ground_gMagnitude and direction angle to water velocity;

step 4.2.6: according to L_KSegment to ground speed V_gAnd L_KCalculating the path length of the segment to obtain the Kth path sub-segment L_KPath time consumption according to L_KCalculating the oil consumption speed of the ship engine of the section, and calculating the Kth path subsection L_KOil consumption of (d);

step 4.3: and obtaining a total oil consumption predicted value of the ship driving to the terminal point of the expected route through integral accumulation.

The present invention may further comprise:

said step 3.1 selecting matrix V_w＝V_w(φ_w,H_w,T_w,N_D) Medium desired speed N_DFor data of the same value, for wave height value H_wWave angle phi relative to the ship's own system_wWave period value T_wThe method for performing three-dimensional linear interpolation specifically comprises the following steps:

independent variables x, y and z are wave height, wave direction and wave period, 3 independent variables phi_w,H_w,T_wAnd a constant value variable N_D0Function V formed_w(φ_w,H_w,T_w,N_D0) The ship is subjected to water stable speed; taking 4 points on the wave height and wave direction plane to form a rectangle A₁A₂A₃A₄Translating out of another rectangle B in the direction of the wave period₁B₂B₃B₄The function values of 8 vertexes of a cuboid formed by two rectangles are determined, and the interpolation work of the independent variable of wave height, wave direction and wave period can be completed by determining the function value of any point in the cuboid;

rectangular parallelepiped A₁A₂A₃A₄-B₁B₂B₃B₄The function value of each vertex in the system is known and is respectively:

V_w(A₁)＝V_w(φ_w1,H_w1,T_w1,N_D0)

V_w(A₂)＝V_w(φ_w2,H_w1,T_w1,N_D0)

V_w(A₃)＝V_w(φ_w1,H_w2,T_w1,N_D0)

V_w(A₄)＝V_w(φ_w2,H_w2,T_w1,N_D0)

V_w(B₁)＝V_w(φ_w1,H_w1,T_w2,N_D0)

V_w(B₂)＝V_w(φ_w2,H_w1,T_w2,N_D0)

V_w(B₃)＝V_w(φ_w1,H_w2,T_w2,N_D0)

V_w(B₄)＝V_w(φ_w2,H_w2,T_w2,N_D0)

then the cuboid A₁A₂A₃A₄-B₁B₂B₃B₄Function value V of any point C_w(C) Comprises the following steps:

the invention has the beneficial effects that:

the invention aims to predict the ship energy consumption by using a data driving method under the condition of no accurate ship energy consumption mathematical model. The invention takes a set of data matrix as support to realize the rapid calculation of the energy consumption of the long navigation path of the ship. The method does not influence the idea of the original ship dynamics simulation model, coexists with the dynamics simulation mode in the actual simulation verification and evaluation system, and is respectively responsible for different types of simulation calculation.

Drawings

Fig. 1 is a schematic diagram of high-dimensional linear interpolation.

Fig. 2 is a block flow diagram of the present invention.

FIG. 3 is a schematic illustration of a desired flight path.

FIG. 4 is a schematic representation of waypoint interpolation.

FIG. 5 is a schematic diagram of the calculation of speed and heading.

Fig. 6 is a basic display effect diagram of the electronic chart.

Fig. 7 is a wave height display effect diagram.

FIG. 8 is a display of predicted values of energy consumption for five routes.

Detailed Description

The invention is further described below with reference to the accompanying drawings.

The invention relates to a ship energy consumption prediction method, in particular to a ship energy consumption prediction method based on data driving, and belongs to the field of ship energy consumption. The invention aims to predict the ship energy consumption by using a data driving method under the condition of not accurately using a ship energy consumption mathematical model.

The invention takes a set of data matrix as support to realize the rapid calculation of the energy consumption of the long navigation path of the ship.

A ship energy consumption prediction method based on data driving comprises the following steps:

step 1: acquiring expected route and departure time T of ship₀And sea state information;

step 2: constructing an off-line basic model, which comprises a ship water steady state navigational speed model, an engine power model and an engine oil consumption model;

the ship water-alignment steady state navigational speed model is composed of a 4-dimensional matrix V_w(φ_w,H_w,T_w,N_D) Represents; v_wRepresenting the steady state navigational speed of the ship to water; phi is a_wRepresenting the wave angle relative to the vessel's own system; h_wRepresents a wave height value; t is_wRepresents the wave period value; n is a radical of_DIs the desired speed of the engine; the engine power model is composed of a 4-dimensional matrix P_m(φ_w,H_w,T_w,V_w) Represents; p_mRepresenting engine power; the engine oil consumption model is composed of a 1-dimensional matrix O_m(P_m) Represents; o is_mRepresenting engine fuel consumption rate;

the three models need to acquire data through a ship pool test or a real ship test, relationship data between respective variables and function values are represented in a normalized grid form, each dimension of matrix input data is sampled into a plurality of points at certain intervals, but a plurality of dimensions are orthogonal to each other to form normalized grid data;

and step 3: carrying out high-dimensional linear interpolation on discrete grid matrixes of the ship to water steady-state navigational speed model and the engine power model; performing one-dimensional linear interpolation on a discrete grid matrix of the engine oil consumption model;

the ship water steady-state navigational speed model and the engine power model are in a 4-dimensional form, and the high-dimensional linear interpolation principle is consistent;

the method for performing high-dimensional linear interpolation on the water steady-state navigational speed model of the ship specifically comprises the following steps:

step 3.1: selection matrix V_w＝V_w(φ_w,H_w,T_w,N_D) Medium desired speed N_DFor data of the same value, for wave height value H_wWave angle phi relative to the ship's own system_wWave period value T_wCarrying out three-dimensional linear interpolation;

independent variables x, y and z are wave height, wave direction and wave period, 3 independent variables phi_w,H_w,T_wAnd a constant value variable N_D0Function V formed_w(φ_w,H_w,T_w,N_D0) The ship is subjected to water stable speed; taking 4 points on the wave height and wave direction plane to form a rectangle A₁A₂A₃A₄Translating out of another rectangle B in the direction of the wave period₁B₂B₃B₄The function values of 8 vertexes of a cuboid formed by two rectangles are determined, and the interpolation work of the independent variable of wave height, wave direction and wave period can be completed by determining the function value of any point in the cuboid;

rectangular parallelepiped A₁A₂A₃A₄-B₁B₂B₃B₄The function value of each vertex in the system is known and is respectively:

V_w(A₁)＝V_w(φ_w1,H_w1,T_w1,N_D0)

V_w(A₂)＝V_w(φ_w2,H_w1,T_w1,N_D0)

V_w(A₃)＝V_w(φ_w1,H_w2,T_w1,N_D0)

V_w(A₄)＝V_w(φ_w2,H_w2,T_w1,N_D0)

V_w(B₁)＝V_w(φ_w1,H_w1,T_w2,N_D0)

V_w(B₂)＝V_w(φ_w2,H_w1,T_w2,N_D0)

V_w(B₃)＝V_w(φ_w1,H_w2,T_w2,N_D0)

V_w(B₄)＝V_w(φ_w2,H_w2,T_w2,N_D0)

then the cuboid A₁A₂A₃A₄-B₁B₂B₃B₄Function value V of any point C_w(C) Comprises the following steps:

step 3.2: the ship water steady state navigational speed model obtained after the three-dimensional linear interpolation and the expected rotating speed N of the engine_DCarrying out two-dimensional linear interpolation to obtain a high-dimensional linear interpolation ship water-steady state navigational speed linear model, and calculating a ship water-steady state navigational speed value under the condition that the wave direction angle, the wave height and the wave period of any ship body system and the expected rotating speed of an engine are known;

and 4, step 4: deducing a ship energy consumption model in real time;

step 4.1: interpolating the expected ship route to form a high-density path point row, ensuring that the maximum distance of each path point is within a set threshold value, wherein the generated path point row contains N path points and N-1 path subsections; calculating longitude and latitude coordinates of each interpolation point in a one-dimensional linear interpolation mode between two route points according to the known route point information, and distributing the known expected engine speed information for each interpolation route point;

step 4.2: traversing each path subsection, and calculating the oil consumption of each path subsection;

step 4.2.1: according to the K-1 th path sub-section L_K-1Time of arrival t_LK-1And L_K-1Acquiring the K path subsection L according to the longitude and latitude information of the end point position of the section_KSea state information of (1), including wave height H_wWave direction and wave period T_w，K＝1,2，...，N-1；

Step 4.2.2: according to the Kth path sub-section L_KCalculating a two-point connecting line vector by utilizing the longitude and latitude information of the starting point and the end point, and calculating the path length and the direction angle of the path vector, wherein the angle is used as the direction angle of the ground speed of the ship;

step 4.2.3: according to L_KDesired engine speed N of the segment_LK-1And the acquired wave height, wave direction and wave period data are substituted into a water steady state navigational speed model V of the ship_w＝V_w(φ_w,H_w,T_w,N_D) Calculating the water speed V of the ship_w；

It is necessary to mix L_KThe wave direction of the segment is transformed into a wave direction angle phi relative to the ship's own train_wReplacing the direction angle of the ship to the water speed with the direction angle of the ship to the ground speed during conversion;

step 4.2.4: will ship to water speed V_wSubstituting sea state information into engine power model P_m＝P_m(φ_w,H_w,T_w,V_w) Calculate L_KThe power of the section engine is substituted into the engine oil consumption model O_m＝O_m(P_m) Calculating the oil consumption speed of the ship engine;

step 4.2.5: according to known L_KSpeed and direction angle L of section-to-ground navigation_KSegment pair water velocity, flow velocity and flow velocity direction angle, and calculating L by using trigonometric relation sine theorem_KSpeed V of section to ground_gMagnitude and direction angle to water velocity;

step 4.2.6: according to L_KSegment to ground speed V_gAnd L_KCalculating the path length of the segment to obtain the Kth path sub-segment L_KPath time consumption according to L_KCalculating the oil consumption speed of the ship engine of the section, and calculating the Kth path subsection L_KOil consumption of (d);

step 4.3: and obtaining a total oil consumption predicted value of the ship driving to the terminal point of the expected route through integral accumulation.

Compared with the prior art, the invention has the beneficial effects that: the energy consumption prediction method provided by the invention takes a set of data matrix as support, and realizes the rapid calculation of energy consumption in a long-route (the navigation time reaches more than several days) mode; in the invention, each item of data in the data matrix can be obtained by numerical calculation, pool experiment or real ship experiment, or multi-channel comprehensive data, and any channel can be directly substituted into the method for calculation as long as the channel meets the requirement of the matrix format, thereby facilitating the subsequent upgrading; the method does not influence the idea of the original ship dynamics simulation model, coexists with the dynamics simulation mode in the actual simulation verification and evaluation system, and is respectively responsible for different types of simulation calculation.

Example 1:

the flow chart of the invention is shown in figure 2, firstly, a desired route is given, the route is composed of a plurality of arbitrary route points, each route point contains longitude and latitude information, the desired rotating speed data of an engine of each route subsection needs to be set, in addition, the sea condition information of different time of the whole navigation area, including wave height, wave direction, wave period and other information, and the departure time of a ship is designated for carrying out time synchronization with the sea condition information; then establishing a high-dimensional matrix of the water steady-state navigational speed of the ship, a high-dimensional matrix of the engine power and an engine oil consumption model so as to construct an offline basic model; completing the linearization operation of the model by a series of matrixes of the offline basic model in a high-dimensional linear interpolation mode; the real-time deduction of the energy consumption model is completed through the set expected air route and the linearization model; and finally, on the basis of the chart core, finishing the visual design work by utilizing a two-dimensional graph display mode.

1. Setting expected route, departure time and sea condition information

The invention needs to complete the following steps:

step 1: the expected course of the ship is determined, the course is composed of course sections formed by connecting a plurality of course points, the coordinates of each course point are a pair of longitude and latitude coordinates, and a plan view is shown in figure 3.

Step 2: determining the time T of track point 1 (start point)₀。

Step 3: obtaining sea condition information including wave height, wave direction and wave period in a period of time after the ship starts.

2. Constructing an offline base model

The off-line basic model in the patent comprises three parts, namely a ship water steady-state navigational speed model, an engine power model and an engine oil consumption model, and how to construct the three models off-line is described as follows:

(1) offline construction of ship-to-water steady-state navigational speed model

The design ship is represented by a 4-dimensional matrix, and the matrix form is as follows:

V_w＝V_w(φ_w,H_w,T_w,N_D)

in the above formula: v_wIndicating the steady state speed of the ship to water, phi_wIndicating the wave angle, H, relative to the vessel's own system_wIndicating the wave height value, T_wRepresents the wave period value, N_DIs the desired speed of the engine.

To obtain a high dimensional matrix as shown in the above formula, the data can be filled into the grid shown in the following table by pool test or real ship test:

TABLE 1 high-dimensional matrix table of ship water-steady-state navigational speed model

According to the form of the table, each row represents a working condition, a water tank test or a real ship test is utilized to obtain the steady-state speed of the ship under the conditions of different wave directions, wave heights, wave periods and expected engine rotating speeds, and the matrix V is completed_w＝V_w(φ_w,H_w,T_w,N_D) And (4) constructing.

(2) Offline construction of a high dimensional matrix of engine power

Designing a ship engine power model represented by a 4-dimensional matrix, the matrix form is as follows:

P_m＝P_m(φ_w,H_w,T_w,V_w)

in the above formula: p_mRepresenting engine power, V_wRepresenting the water speed of the ship, and the rest variables are consistent with those in (1).

To obtain a high dimensional matrix as shown in the above formula, the data can be filled into the grid shown in the following table by pool test or real ship test:

TABLE 2 Ship engine power model high-dimensional matrix table

According to the form of the table, each row represents a working condition, and engine power under the conditions of different wave directions, wave heights, wave periods and ship water speed is obtained by utilizing a pool test or a real ship test to complete the matrix P_m＝P_m(φ_w,H_w,T_w,V_w) And (4) constructing.

(3) Offline construction of engine oil consumption model

Designing the oil consumption model of the ship engine is represented by a 1-dimensional matrix, wherein the matrix form is as follows:

O_m＝O_m(P_m)

in the above formula: o is_mIndicating engine oil consumption, P_mRepresenting engine power.

To obtain a high dimensional matrix as shown in the above formula, the data can be filled into the grid shown in the following table by pool test or real ship test:

TABLE 3 high-dimensional matrix table of oil consumption model of marine engine

Oil consumption of engine Om	Engine power Pm
		O1	P1
O2	P2
		O3	P3
……	……

According to the form of the table, each row represents a working condition, and the engine oil consumption under different engine power conditions is obtained by using a pool test or a real ship test to complete the matrix O_m＝O_m(P_m) And (4) constructing.

Note that the data in the above table is normalized grid data, each dimension of the matrix input data is sampled into a plurality of points at certain intervals, but a plurality of dimensions are orthogonal to each other, and the normalized grid data is formed.

3. High-dimensional linear interpolation of model matrices

The method is characterized in that linear interpolation is carried out on a discrete grid matrix of a water steady-state navigational speed model, an engine power model and an engine oil consumption model of a ship, because the water steady-state navigational speed model and the engine power model are in a 4-dimensional form, the high-dimensional linear interpolation principle is consistent, the water steady-state navigational speed model is described as an example, and the interpolation method is shown in figure 1.

Step 1: selecting a desired speed N of the engine_DFor the same value, i.e. selecting N in Table 1_DAnd performing three-dimensional linear interpolation on wave height, wave direction and wave period by using the data of the row with the same value.

Independent variables x, y and z are wave height, wave direction and wave period, 3 independent variables phi_w,H_w,T_wAnd a constant value variable N_D0Function V of formation_w(φ_w,H_w,T_w,N_D0) And the ship is at the water stable speed. Taking 4 points on the wave height and wave direction plane to form a rectangle A₁A₂A₃A₄Translating out of another rectangle B in the direction of the wave period₁B₂B₃B₄The function values of 8 fixed points of a cuboid formed by two rectangles are determined, and the interpolation work of the independent variable of wave height, wave direction and wave period can be completed by determining the function value of any point in the cuboid. The whole idea is that A₁、A₂And A₃、A₄Determination of A₅、A₆、A₇Point function value of from₁、B₂And B₃、B₄Determination of B₅、B₆、B₇Point function value of from₇And B₇And determining the value of the C point function. In a rectangular parallelepiped A₁A₂A₃A₄-B₁B₂B₃B₄The function value of any point C is as follows:

in the above formula:

V_w(A₁)＝V_w(φ_w1,H_w1,T_w1,N_D0)

V_w(A₂)＝V_w(φ_w2,H_w1,T_w1,N_D0)

V_w(A₃)＝V_w(φ_w1,H_w2,T_w1,N_D0)

V_w(A₄)＝V_w(φ_w2,H_w2,T_w1,N_D0)

V_w(B₁)＝V_w(φ_w1,H_w1,T_w2,N_D0)

V_w(B₂)＝V_w(φ_w2,H_w1,T_w2,N_D0)

V_w(B₃)＝V_w(φ_w1,H_w2,T_w2,N_D0)

V_w(B₄)＝V_w(φ_w2,H_w2,T_w2,N_D0)

step 2: v obtained by high-dimensional linear interpolation in Step 1_wcWith desired speed N of the engine_DTwo-dimensional linear interpolation is carried out, the interpolation principle is similar to that of Step 1, but the interpolation is carried out only in the x plane and the y plane, so that the detailed description is omitted here.

Through the operation of the Step 1 and the Step 2, a high-dimensional linear interpolation ship-to-water steady-state navigational speed linear model can be obtained, namely, the ship-to-water steady-state navigational speed value can be calculated by knowing the wave direction angle, the wave height and the wave period of any ship body system and the expected rotational speed of the engine, and the engine power linear model can also be obtained by adopting the same high-dimensional linear interpolation principle.

The engine oil consumption model in the patent only contains one variable and can be obtained by adopting a one-dimensional linear interpolation mode, and the description is not provided.

4. Real-time deduction ship energy consumption model

This patent will realize the deduction of boats and ships energy consumption model through following step:

step 1: for a specified route, interpolating the route to form a high-density route point row, wherein the interpolation effect is shown in fig. 4, the step ensures that the maximum distance between each route point is within a certain threshold value, the threshold value is defined to be 500m, according to the known route point information, the longitude and latitude coordinates of each interpolation point are easily obtained through a one-dimensional linear interpolation mode between two route points, and the known expected engine speed information is distributed to each interpolation route point.

Step 2: assuming that the number of newly generated waypoints is N, the newly generated waypoints totally comprise N-1 path subsections, and then each path subsection is traversed, the patent takes the current section as the subsection between the waypoints K and K +1 as an example for explanation, and the waypoint is called L here_K. According to L_K-1Segment calculated arrival time

And L_K-1The longitude and latitude information of the end position of the segment, and the sea condition information of the time and the space position, including wave height, wave direction and wave period, are obtained and used as the whole L_KAverage sea state value of the segment.

Step 3: according to L_KAnd calculating a two-point connecting line vector by utilizing the longitude and latitude information of the starting point and the ending point of the section through geographic coordinate transformation, and then calculating the path length (great circle distance) and the direction angle of the path vector, wherein the angle is used as the direction angle of the ground speed of the ship.

Step 4: according to L_KDesired engine speed of the segment

And substituting the wave height, wave direction and wave period data acquired in Step 2 into V_w＝V_w(φ_w,H_w,T_w,N_D) Model, calculating the water speed of ship, and paying attention to the wave direction phi_wThe conversion to the system requires that the direction angle of the ship to the water speed is known, the value is unknown, and the direction angle of the ship to the ground speed is approximately substituted for the direction angle of the ship to the ground speed for simplifying the processing.

Step 5: v calculated in Step 4_wAnd sea state information, substituting into P_m＝P_m(φ_w,H_w,T_w,V_w) Model, calculate L_KThe engine power of the segment is then substituted into O_m＝O_m(P_m) And calculating the oil consumption speed of the ship engine.

Step 6: according to the currently known L_KSpeed and direction angle L of section-to-ground navigation_KSegment pair water velocity, flow velocity and flow velocity direction angle, and calculating L by using trigonometric relation sine theorem_KThe section is relative to the ground speed and the direction angle of the water speed, the principle is shown in figure 5, wherein the water speed V_wHas solved the magnitude of (1), has unknown direction, and has a ground speed V_gIs known in direction, unknown in magnitude, magnitude and direction V of the flow velocity_fAre all known, at which time the velocity vectorTheta in triangle_fgCan solve the ground speed V by using the cosine law_gTo obtain the water velocity V_wIn the direction of (a).

Step 7: according to L_KCalculating the path time consumption of the subsection according to the speed of the subsection to the ground and the path length of the subsection calculated in Step 3, and calculating L through integral accumulation_KThe total time consumption after the section is ended, the time consumption of the section and the oil consumption speed calculated in Step 6 are utilized to calculate the oil consumption quality of the section, and the L is calculated through integral accumulation_KTotal oil consumption after the end of the session.

5. Visualization design of energy consumption deduction

The method is based on the electronic chart core and used for designing visualization of ship energy consumption deduction. The electronic chart core is a series of complex data structures, according to which the patent displays the basic map contour of the world and most of the land and island contours, and the display effect of the electronic chart is shown in fig. 6.

The electronic chart is displayed under the Kyork coordinate system, so that the coordinates of each pixel point can be calculated, expected course and route point information is known in the patent, so that the expected course and route point information can be displayed under the background of the chart, and the sea state information takes longitude and latitude and time as independent variables.

In the patent, the principle of course information display is that adjacent route points are directly connected by line segments, and for sea state information, the wave height is taken as an example, and the maximum value of the wave height is set as H_maxWhen the actual value of the wave height at a certain latitude is H_actThen, color value C is displayed on the electronic chart_lanRepresented by the formula:

the above formula has the following three cases:

(1) if 0 is less than or equal to C_lan< 255, the RGB pixel at the longitude and latitude point is (C)_lan，0，0)；

(2) If 255 is less than or equal to C_lan＜511, RGB pixel at the longitude and latitude point is (0, C)_lan-255，0)；

(3) If 512 is less than or equal to C_lanLess than or equal to 768, the RGB pixel at the longitude and latitude point is (0, 0, C)_lan-255)。

The principle of displaying other sea state information, including the wave direction and the wave period, by colors is the same as the wave height, and the patent does not describe here, and the two-dimensional display effect of the wave height is shown in fig. 7.

In order to increase the display effect of the predicted value of the energy consumption of the ship, the invention also utilizes different colors to display the energy consumption service condition of the ship on an expected navigation line, and the distribution of the different colors is the same as the color distribution principle of the wave height. FIG. 8 shows 5 expected routes, and after the departure time is set and the sea state information is loaded, the display effect of the predicted value of the ship energy consumption is obtained by using the calculation method of the patent.

The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention, and various modifications and changes may be made by those skilled in the art. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims (2)

1. A ship energy consumption prediction method based on data driving is characterized by comprising the following steps:

step 1: acquiring expected route and departure time T of ship₀And sea state information;

step 2: constructing an off-line basic model, which comprises a ship water steady state navigational speed model, an engine power model and an engine oil consumption model;

the ship water-alignment steady state navigational speed model is composed of a 4-dimensional matrix V_w(φ_w,H_w,T_w,N_D) Represents; v_wRepresenting the steady state navigational speed of the ship to water; phi is a_wRepresenting the wave angle relative to the vessel's own system; h_wRepresents a wave height value; t is_wRepresents the wave period value; n is a radical of_DIs the desired speed of the engine; the engine power model is composed of a 4-dimensional matrix P_m(φ_w,H_w,T_w,V_w) Represents; p_mRepresenting engine power; the engine oil consumption model is composed of a 1-dimensional matrix O_m(P_m) Represents; o is_mRepresenting engine fuel consumption rate;

the three models need to acquire data through a ship pool test or a real ship test, the relation data between respective variables and function values are expressed in a normalized grid form, each dimension of matrix input data is sampled into a plurality of points at certain intervals, but the plurality of dimensions are orthogonal to each other to form normalized grid data;

and step 3: carrying out high-dimensional linear interpolation on discrete grid matrixes of the ship to water steady-state navigational speed model and the engine power model; performing one-dimensional linear interpolation on a discrete grid matrix of the engine oil consumption model;

the ship water steady-state navigational speed model and the engine power model are in a 4-dimensional form, and the high-dimensional linear interpolation principle is consistent;

the method for performing high-dimensional linear interpolation on the water steady-state navigational speed model of the ship specifically comprises the following steps:

step 3.1: selection matrix V_w＝V_w(φ_w,H_w,T_w,N_D) Medium desired speed N_DFor data of the same value, for wave height value H_wWave angle phi relative to the ship's own system_wWave period value T_wCarrying out three-dimensional linear interpolation;

step 3.2: the ship water steady state navigational speed model obtained after the three-dimensional linear interpolation and the expected rotating speed N of the engine_DCarrying out two-dimensional linear interpolation to obtain a high-dimensional linear interpolation ship water-steady state navigational speed linear model, and calculating a ship water-steady state navigational speed value under the condition that the wave direction angle, the wave height and the wave period of any ship body system and the expected rotating speed of an engine are known;

and 4, step 4: deducing a ship energy consumption model in real time;

step 4.1: for an expected course of a ship, interpolating the expected course to form a high-density path point row, ensuring that the maximum distance between all path points is within a set threshold value, wherein the generated path point row contains N path points and N-1 path subsections; according to the known waypoint information, calculating the longitude and latitude coordinates of each interpolation point in a one-dimensional linear interpolation mode between two waypoints, and distributing the known expected engine speed information for each interpolation waypoint;

step 4.2: traversing each path subsection, and calculating the oil consumption of each path subsection;

step 4.2.1: according to the K-1 th path sub-section L_K-1Time of arrival of

And L_K-1Acquiring the K-th path subsection L according to the longitude and latitude information of the end point position of the section_KSea state information of (1), including wave height H_wWave direction and wave period T_w，K＝1,2，...，N-1；

Step 4.2.2: according to the Kth path sub-section L_KCalculating a two-point connecting line vector by utilizing the longitude and latitude information of the starting point and the end point, and calculating the path length and the direction angle of the path vector, wherein the angle is used as the direction angle of the ground speed of the ship;

step 4.2.3: according to L_KDesired engine speed of the segment

And substituting the acquired wave height, wave direction and wave period data into a ship water steady state navigational speed model V_w＝V_w(φ_w,H_w,T_w,N_D) Calculating the water speed V of the ship_w；

It is necessary to mix L_KThe wave direction of the segment is transformed into a wave direction angle phi relative to the ship's own train_wReplacing the direction angle of the ship to the water speed with the direction angle of the ship to the ground speed during conversion;

step 4.2.4: will ship to water speed V_wSubstituting sea state information into engine power model P_m＝P_m(φ_w,H_w,T_w,V_w) Calculate L_KThe engine power of the section is then substituted into the engineOil consumption model O_m＝O_m(P_m) Calculating the oil consumption speed of the ship engine;

step 4.2.5: according to known L_KSpeed and direction angle L of section-to-ground navigation_KSegment pair water velocity, flow velocity and flow velocity direction angle, and calculating L by using trigonometric relation sine theorem_KSpeed V of section to ground_gMagnitude and direction angle to water velocity;

step 4.2.6: according to L_KSegment to ground speed V_gAnd L_KCalculating the path length of the segment to obtain the Kth path sub-segment L_KPath time consumption according to L_KCalculating the oil consumption speed of the ship engine of the section, and calculating the Kth path subsection L_KOil consumption of (d);

step 4.3: and obtaining a total oil consumption predicted value of the ship driving to the terminal point of the expected route through integral accumulation.

2. The method for predicting the energy consumption of the ship based on data driving according to claim 1, wherein the method comprises the following steps: said step 3.1 selecting matrix V_w＝V_w(φ_w,H_w,T_w,N_D) Medium desired speed N_DFor data of the same value, for wave height value H_wWave angle phi relative to the ship's own system_wWave period value T_wThe method for performing three-dimensional linear interpolation specifically comprises the following steps:

independent variables x, y and z are wave height, wave direction and wave period, 3 independent variables phi_w,H_w,T_wAnd a constant value variable N_D0Function V of formation_w(φ_w,H_w,T_w,N_D0) The ship is subjected to water stable speed; taking 4 points on the wave height and wave direction plane to form a rectangle A₁A₂A₃A₄Translating out of another rectangle B in the direction of the wave period₁B₂B₃B₄The function values of 8 vertexes of a cuboid formed by two rectangles are determined, and the interpolation work of the independent variable of wave height, wave direction and wave period can be completed by determining the function value of any point in the cuboid;

rectangular parallelepiped A₁A₂A₃A₄-B₁B₂B₃B₄The function value of each vertex in the system is known and is respectively:

V_w(A₁)＝V_w(φ_w1,H_w1,T_w1,N_D0)

V_w(A₂)＝V_w(φ_w2,H_w1,T_w1,N_D0)

V_w(A₃)＝V_w(φ_w1,H_w2,T_w1,N_D0)

V_w(A₄)＝V_w(φ_w2,H_w2,T_w1,N_D0)

V_w(B₁)＝V_w(φ_w1,H_w1,T_w2,N_D0)

V_w(B₂)＝V_w(φ_w2,H_w1,T_w2,N_D0)

V_w(B₃)＝V_w(φ_w1,H_w2,T_w2,N_D0)

V_w(B₄)＝V_w(φ_w2,H_w2,T_w2,N_D0)

then the cuboid A₁A₂A₃A₄-B₁B₂B₃B₄Function value V of any point C_w(C) Comprises the following steps:

Images