Ocean vessel oil consumption prediction method based on big data

Abstract

The invention provides an ocean vessel oil consumption prediction method based on big data, and belongs to the field of data prediction. The problem of complex data prediction is solved; the method specifically comprises the following steps: S1, acquiring ship data and climate data, and acquiring oil consumption information; s2, processing the ship data and the climate data, solving a correlation coefficient in combination with the oil consumption information, and analyzing and judging the correlation coefficient to obtain an influence parameter; s3, according to the influence parameters, the acting force of the influence parameters on the ship is obtained, the acting force is integrated, propulsive force is obtained, regression analysis is conducted according to the propulsive force in combination with oil consumption information, and an oil consumption prediction model is constructed; s4, predicting the oil consumption of the ship according to the oil consumption prediction model, performing real-time analysis on the oil consumption of the ship, and performing early warning when the oil consumption is abnormal; by predicting the oil consumption of the ocean-going ship, data reference is provided for the ocean-going ship, and safe operation of the ocean-going ship is guaranteed.

Description

Technical Field

[0001] The present invention discloses a method for predicting fuel consumption of ocean-going ships based on big data, and relates to the field of data prediction. Background Art

[0002] The existing methods for predicting fuel consumption of ocean-going ships have the following shortcomings:

[0003] Incomplete data: The amount of data generated by ocean-going ships during navigation is huge and complex, including multiple factors such as navigation speed, location, climate, load, etc.; existing invention data is often incomplete, which brings challenges to fuel consumption prediction;

[0004] Insufficient data preprocessing: Existing inventions often do not preprocess data sufficiently, which affects the accuracy and stability of subsequent prediction models;

[0005] Excessive model simplification: Existing fuel consumption prediction models are often based on some simplified assumptions and formulas, such as the speed-power model and the thrust-power model. Although these models can reflect the relationship between fuel consumption and certain factors to a certain extent, they often ignore the influence of other important factors, such as the specific design of the ship and changes in climatic conditions. Summary of the invention

[0006] In view of the shortcomings of the prior art, the purpose of the present invention is to provide a method for predicting fuel consumption of ocean-going ships based on big data, aiming to solve the problem of complex data prediction.

[0007] In order to achieve the above object, the present invention is implemented by the following technical solution: a method for predicting fuel consumption of ocean-going ships based on big data, the prediction method comprising:

[0008] Step S1: Obtain ship data and climate data, and obtain fuel consumption information;

[0009] Step S2: Obtain ship parameters and climate parameters according to ship data and climate data, process the ship parameters and climate parameters, extract abnormal values ​​in the ship parameters and climate parameters, calculate the correlation coefficient between the ship parameters and climate parameters in combination with fuel consumption information, analyze and judge the correlation coefficient, and obtain the influencing parameters;

[0010] Step S3: Obtain the force of the influencing parameters on the ship, perform force analysis on the force on the ship, obtain the total force, perform regression analysis based on the total force combined with the fuel consumption information, and build a fuel consumption prediction model;

[0011] Step S4: Perform navigation prediction and real-time prediction of the fuel consumption of the ship according to the fuel consumption prediction model, and issue an alarm for abnormal fuel consumption generated during the navigation process.

[0012] Furthermore, the specific steps of step S2 are as follows:

[0013] Step S21: Processing the ship data and climate data to obtain ship parameters and climate parameters, performing numerical analysis on the ship parameters and climate parameters to obtain the distribution of the ship parameters and climate parameters, and extracting abnormal values ​​in the ship parameters and climate parameters according to the distribution;

[0014] Step S22: according to the processed data, the mean of the ship parameters and the climate parameters is obtained, and the correlation coefficient between the ship parameters, the climate parameters and the fuel consumption is calculated in combination with the fuel consumption information;

[0015] Step S23: Analyze and judge the correlation coefficient, extract the ship parameters and climate parameters with large correlation coefficients, and obtain the influencing parameters.

[0016] Furthermore, the specific steps of step S21 are as follows:

[0017] Step S211: Obtain ship parameter type y, obtain data of x ships according to ship data, and obtain ship parameter cb(i, cy), 0<cy≤y, cy refers to the specific ship parameter type; obtain the climate type z to which the ship is subject, obtain the climate impact on x ships according to the climate data, and obtain climate parameter qh(i, qz), 0<qz≤z, qz refers to the specific ship parameter type, and i refers to the i-th ship;

[0018] Step S212: Perform numerical analysis on the ship parameter cb(i, cy) and the climate parameter qh(i, qz), and determine whether the parameters conform to the normal distribution according to the analysis results;

[0019] Step S213: extracting abnormal values ​​according to the judgment result;

[0020] If the parameters conform to the normal distribution, outliers are extracted according to the 3σ principle;

[0021] If the parameters do not conform to the normal distribution, outliers are extracted according to the isolation forest method.

[0022] Furthermore, the specific steps of step S212 are as follows:

[0023] Step S2121: Calculate the mean of the ship parameters to obtain the mean μ(cy) of the ship parameters;

[0024]

[0025] Step S2122: According to the ship parameters, combined with the mean of the ship parameters, the standard deviation of the ship parameters is obtained to obtain the ship parameter standard deviation σ(cy):

[0026]

[0027] Step S2123: According to the mean value μ(cy) of the ship parameters, the standard deviation σ(cy) of the ship parameters, and in combination with the ship parameters, the kurtosis and skewness of the ship parameters are obtained to obtain the ship parameter kurtosis FD(cy) and the ship parameter skewness PD(cy):

[0028]

[0029] Step S2124: Analyze and judge the ship parameters according to the ship parameter kurtosis FD(cy) and the ship parameter skewness PD(cy):

[0030] If -2≤FD(cy)≤2, and -0.5≤PD(cy)≤0.5, the ship parameters conform to the normal distribution;

[0031] If FD(cy)>2 or FD(cy)<-2 or PD(cy)>0.5 or PD(cy)<-0.5, the ship parameters do not conform to the normal distribution.

[0032] Furthermore, the step S212 further includes:

[0033] Step S2125: Calculate the mean of the climate parameter to obtain the climate parameter mean μ(qz);

[0034] Step S2126: according to the climate parameter and in combination with the mean value of the climate parameter, the standard deviation of the climate parameter is obtained to obtain the climate parameter standard deviation σ(qz);

[0035] Step S2127: according to the climate parameter mean μ(qz), the climate parameter standard deviation σ(qz), and the climate parameter, the kurtosis and skewness of the climate parameter are calculated in combination with the climate parameter to obtain the climate parameter kurtosis FD(qz) and the climate parameter skewness PD(qz);

[0036] By obtaining the kurtosis and skewness of the climate parameters, the data distribution of the climate parameters is judged, and the parameters that conform to the normal distribution are extracted through the kurtosis and skewness data;

[0037] Step S2128: Analyze and judge the climate parameters according to the climate parameter kurtosis FD(qz) and the climate parameter skewness PD(qz):

[0038] If -2≤FD(qz)≤2, and -0.5≤PD(qz)≤0.5, the climate parameter conforms to the normal distribution;

[0039] If FD(qz)>2 or FD(qz)<-2 or PD(qz)>0.5 or PD(qz)<-0.5, the climate parameter does not conform to the normal distribution.

[0040] Furthermore, the specific steps of step S22 are as follows:

[0041] Step S221: According to the fuel consumption information, the fuel consumption of the ship YH(i, cy, qz) is obtained, and the average fuel consumption of the ship under different climates and the average fuel consumption under different ship data are obtained to obtain the average fuel consumption of the ship CJZ and the average fuel consumption of the climate QJZ;

[0042] The calculation process of the ship fuel consumption mean CJZ is as follows:

[0043]

[0044] YH(i, cy, qz) refers to the fuel consumption of the i-th ship under the cy-th ship parameters and under the qz-th climate influence, x refers to the number of ships, and z refers to the climate type;

[0045] The calculation process of climate fuel consumption mean QJZ is as follows:

[0046]

[0047] Among them, y refers to the type of ship parameters;

[0048] Step S222: Obtain ship parameters cb(i, cy) and climate parameters qh(i, qz), and obtain the mean values ​​of ship parameters J(cy) and climate parameters J(qz);

[0049]

[0050] Step S223: according to the ship parameter cb(i, cy), the mean value of the ship parameter J(cy), the fuel consumption of the ship YH(i, cy, qz), and the mean value of the fuel consumption of the ship CJZ, the correlation between the ship parameter and the fuel consumption of the ship is obtained to obtain the correlation coefficient XG(cy);

[0051]

[0052] Step S224: according to the climate parameter qh(i, qz), the climate parameter mean J(qz), the fuel consumption of the ship YH(i, cy, qz), and the climate fuel consumption mean CJZ, the correlation between the climate parameter and the ship fuel consumption is obtained to obtain the correlation coefficient XG(qz);

[0053]

[0054] Furthermore, the subsequent steps of step S23 are as follows:

[0055] Step S231: Obtain the correlation coefficient threshold t, 0<t≤1, obtain the correlation coefficient XG(cy) between the ship parameters and the ship fuel consumption, and make a judgment and analysis on XG(cy) in combination with the correlation coefficient threshold t:

[0056] If |XG(cy)|≥t, the ship parameter is an influencing parameter and the ship parameter is extracted;

[0057] If |XG(cy)|<t, then the ship parameter does not belong to the influencing parameter;

[0058] Step S232: Obtain the correlation coefficient XG(qz) between the climate parameter and the ship fuel consumption, and make a judgment and analysis on XG(qz) in combination with the correlation coefficient threshold t:

[0059] If |XG(qz)|≥t, the climate parameter is an influencing parameter and the climate parameter is extracted;

[0060] If |XG(qz)|<t, then the climate parameter is not an influencing parameter.

[0061] Furthermore, the specific steps of step S3 are as follows:

[0062] Step S31: Obtain the forces F(i, cy) and F(i, qz) of the influencing parameters on the ship; obtain the angle θ between the force and the ship's travel direction cy ,θ qz , obtain the friction coefficient xs of the hull; integrate the forces to obtain the total force zF i ;

[0063] zF i =F(i,cy)cosθ cy +xs×F(i,cy)sinθ cy +F(i,qz)cosθ qz +xs×F(i,qz)sinθ qz ;

[0064] Step S32: Obtain the fuel consumption xh of the ship according to the fuel consumption information i , combined with the total force, construct the total force zF i Fuel consumption xh i The data set is processed by least squares to obtain the first parameter k and the second parameter v of the fuel consumption prediction model;

[0065] According to the first parameter k and the second parameter v, a fuel consumption prediction model is constructed to obtain a fuel consumption prediction model. The fuel consumption prediction model is as follows:

[0066] xh=k×zF+v

[0067] xh is the fuel consumption variable of the fuel consumption prediction model, and zF is the total force variable of the fuel consumption prediction model.

[0068] Compared with the prior art, the present invention has the following beneficial effects:

[0069] Improve the data scope: The present invention obtains hull data through sensors and recording equipment, obtains climate data through meteorological bureaus, ocean observation stations or professional meteorological service agencies, and analyzes and judges hull data and climate data;

[0070] Strengthen data preprocessing: The present invention judges each type of data through numerical analysis, and extracts outliers using the 3σ principle or isolation forest method according to the judgment results;

[0071] Optimizing model parameters: The present invention obtains the correlation between ship parameters and fuel consumption through ship parameters, ship parameter mean values, ship fuel consumption, and ship fuel consumption mean values. It also obtains the correlation between climate parameters and fuel consumption through climate parameters, climate parameter mean values, ship fuel consumption, and ship fuel consumption mean values. According to the size of the correlation value, the main parameters of fuel consumption are selected, and irrelevant parameters are excluded;

[0072] Improved model building method: The present invention obtains the forces exerted on the ship by various ship parameters and climate parameters during the operation of the ship, performs force analysis on these forces, obtains the forces affecting fuel consumption, and builds a model based on the correlation between the forces and fuel consumption. BRIEF DESCRIPTION OF THE DRAWINGS

[0073] Other features, objects and advantages of the present invention will become more apparent from the detailed description of non-limiting embodiments made with reference to the following drawings:

[0074] Figure 1 It is a schematic diagram of the method of the present invention;

[0075] Figure 2 This is a schematic diagram of the main data processing of the present invention;

[0076] Figure 3 Schematic diagram of regression analysis of the present invention. DETAILED DESCRIPTION

[0077] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and understandable, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0078] Embodiment 1

[0079] See also Figure 1 ,The fuel consumption prediction methods of ocean-going ships based on big data include:

[0080] Step S1: Obtain ship data and climate data, and obtain fuel consumption information;

[0081] Step S11: Obtain ship data based on sensors and recording equipment, including but not limited to the ship's type, tonnage, draft, speed, and heading.

[0082] Step S12: Obtain the latest climate data from the Meteorological Bureau, ocean observation station or professional meteorological service agency; these climate data include wind speed, wind direction, wave height, ocean current direction and speed, temperature and humidity.

[0083] Step S13: Obtain fuel consumption information, including fuel consumption records of the ship in past voyages, fuel consumption performance under different types of navigation tasks, and detailed data on engine working efficiency and fuel consumption rate.

[0084] Step S2: Process the ship data and climate data, obtain the correlation coefficient based on the processed data and the fuel consumption information, analyze and judge the correlation coefficient, and obtain the influencing parameters;

[0085] Step S21: Processing the ship data and climate data to obtain ship parameters and climate parameters, performing numerical analysis on the ship parameters and climate parameters to obtain the distribution of the ship parameters and climate parameters, extracting abnormal values ​​in the ship parameters and climate parameters according to the distribution, and eliminating the influence of abnormal values ​​on the data;

[0086] The specific steps for extracting outliers are as follows:

[0087] Step S211: Obtain ship parameter type y, obtain data of x ships according to ship data, and obtain ship parameter cb(i, cy), 0<cy≤y, where: cy refers to a specific ship parameter type, such as: ship weight, ship size, ship material, ship age, ship speed; obtain the climate type z to which the ship is subject, obtain the climate impact on x ships according to the climate data, and obtain climate parameter qh(i, qz), 0<qz≤z, where: qz refers to a specific ship parameter type, such as: wind force, seawater flow rate, temperature;

[0088] Step S212: numerically analyzing the ship parameter cb(i, cy) and the climate parameter qh(i, qz), and judging whether the parameters conform to the normal distribution according to the analysis results;

[0089] Step S2121: Calculate the mean of the ship parameters to obtain the mean μ(cy) of the ship parameters;

[0090]

[0091] Step S2122: According to the ship parameters, combined with the mean of the ship parameters, the standard deviation of the ship parameters is obtained to obtain the ship parameter standard deviation σ(cy):

[0092]

[0093] Step S2123: According to the mean value μ(cy) of the ship parameters, the standard deviation σ(cy) of the ship parameters, and in combination with the ship parameters, the kurtosis and skewness of the ship parameters are obtained to obtain the ship parameter kurtosis FD(cy) and the ship parameter skewness PD(cy):

[0094]

[0095] In the process of obtaining the kurtosis FD of the ship parameters in this application document, by subtracting 3 from the logarithm, the kurtosis and skewness values ​​are made close to 0. By obtaining the kurtosis and skewness of the ship parameters, the data distribution of the ship parameters is judged, and the parameters that conform to the normal distribution are extracted through the kurtosis and skewness data. This method has high reliability and simplicity.

[0096] Step S2124: Analyze and judge the ship parameters according to the ship parameter kurtosis FD(cy) and the ship parameter skewness PD(cy):

[0097] If -2≤FD(cy)≤2, and -0.5≤PD(cy)≤0.5, the ship parameters conform to the normal distribution;

[0098] If FD(cy)>2 or FD(cy)<-2 or PD(cy)>0.5 or PD(cy)<-0.5, the ship parameters do not conform to the normal distribution;

[0099] Step S2125: Calculate the mean of the climate parameter to obtain the climate parameter mean μ(qz);

[0100]

[0101] Step S2126: According to the climate parameter, combined with the mean value of the climate parameter, the standard deviation of the climate parameter is obtained to obtain the climate parameter standard deviation σ(qz):

[0102]

[0103] Step S2127: According to the climate parameter mean μ(qz), the climate parameter standard deviation σ(qz), and the climate parameter, the kurtosis and skewness of the climate parameter are obtained in combination with the climate parameter to obtain the climate parameter kurtosis FD(qz) and the climate parameter skewness PD(qz):

[0104]

[0105] By obtaining the kurtosis and skewness of climate parameters, the data distribution of climate parameters is judged, and the parameters that conform to the normal distribution are extracted through the kurtosis and skewness data. This method is highly reliable and simple.

[0106] Step S2128: Analyze and judge the climate parameters according to the climate parameter kurtosis FD(qz) and the climate parameter skewness PD(qz):

[0107] If -2≤FD(qz)≤2, and -0.5≤PD(qz)≤0.5, the climate parameter conforms to the normal distribution;

[0108] If FD(qz)>2 or FD(qz)<-2 or PD(qz)>0.5 or PD(qz)<-0.5, the climate parameter does not conform to the normal distribution.

[0109] Step S213: extracting abnormal values ​​according to the judgment result;

[0110] If the parameters conform to the normal distribution, outliers are extracted according to the 3σ principle;

[0111] For example, for ship parameters, obtain the mean μ(cy) and standard deviation σ(cy) of ship parameters and compare the ship parameters:

[0112] If μ(cy)-3σ(cy)≤cb(i,cy)≤μ(cy)+3σ(cy), then cb(i,cy) is a normal value;

[0113] If μ(cy)-3σ(cy)>cb(i,cy) or cb(i,cy)<μ(cy)+3σ(cy), then cb(i,cy) is an outlier;

[0114] If the parameters do not conform to the normal distribution, outliers are extracted according to the isolation forest method.

[0115] It should be noted that Isolation Forest (iForest) is an unsupervised anomaly detection algorithm based on a tree structure. Its core idea is that abnormal data points are more likely to be randomly segmented and quickly isolated because of their small number and large differences from normal data.

[0116] Step S22: according to the processed data, the mean of the ship parameters and the climate parameters is obtained, and the correlation coefficient between the ship parameters, the climate parameters and the fuel consumption is calculated in combination with the fuel consumption information;

[0117] Step S221: According to the fuel consumption information, the fuel consumption of the ship YH(i, cy, qz) is obtained, and the average fuel consumption of the ship under different climates and the average fuel consumption under different ship data are obtained to obtain the average fuel consumption of the ship CJZ and the average fuel consumption of the climate QJZ;

[0118] The calculation process of the ship fuel consumption mean CJZ is as follows:

[0119]

[0120] It should be noted that: YH(i, cy, qz) refers to the fuel consumption of the i-th ship under the cy-th ship parameters and the qz-th climate influence.

[0121] According to the formula for calculating the average fuel consumption CJZ of ships, the fuel consumption of x ships under z types of climate is accumulated, divided by the product of the number of ships and the type of climate, the average fuel consumption under the same ship parameters is calculated, and the impact data of the ship is calculated as a whole, which simplifies the calculation steps;

[0122] The calculation process of climate fuel consumption mean QJZ is as follows:

[0123]

[0124] According to the formula for calculating the average fuel consumption CJZ of ships, the fuel consumption of x ships under y ship parameters is accumulated, divided by the product of the number of ships and the type of ship parameters, the average fuel consumption under the same climate is calculated, and the impact data of the ships is calculated as a whole, which simplifies the calculation steps;

[0125] Step S222: Obtain ship parameters cb(i, cy) and climate parameters qh(i, qz), calculate the mean of the ship parameters to obtain the ship parameter mean J(cy), calculate the mean of the climate parameters to obtain the climate parameter mean J(qz);

[0126]

[0127] Step S223: according to the ship parameter cb(i, cy), the mean value of the ship parameter J(cy), the fuel consumption of the ship YH(i, cy, qz), and the mean value of the fuel consumption of the ship CJZ, the correlation between the ship parameter and the fuel consumption of the ship is obtained to obtain the correlation coefficient XG(cy);

[0128]

[0129] Step S224: according to the climate parameter qh(i, qz), the climate parameter mean J(qz), the fuel consumption of the ship YH(i, cy, qz), and the climate fuel consumption mean CJZ, the correlation between the climate parameter and the ship fuel consumption is obtained to obtain the correlation coefficient XG(qz);

[0130]

[0131] By obtaining the correlation coefficient, the correlation relationship between the parameter and fuel consumption is obtained. The value range of the correlation coefficient is [-1, 1]. If the correlation coefficient is greater than 0, it indicates that the parameter and fuel consumption are positively correlated. If the correlation coefficient is less than 0, it indicates that the parameter and fuel consumption are negatively correlated. If the correlation coefficient is equal to 0, it indicates that there is no correlation between the parameter and fuel consumption. The greater the absolute value of the correlation coefficient, the higher the correlation between the parameter and fuel consumption. The greater the impact of the parameter on fuel consumption, the data can be divided by the correlation coefficient, and the correlation coefficient of a specific numerical range can be selected according to specific requirements.

[0132] Step S23: Analyze and judge the correlation coefficient, extract the ship parameters and climate parameters with large correlation coefficients, and obtain the influencing parameters;

[0133] Step S231: According to actual needs, obtain the correlation coefficient threshold t, obtain the correlation coefficient XG(cy) between the ship parameters and the ship fuel consumption, and make a judgment and analysis on XG(cy) in combination with the correlation coefficient threshold t:

[0134] If |XG(cy)|≥t, the ship parameter is an influencing parameter and the ship parameter is extracted;

[0135] If |XG(cy)|<t, then the ship parameter does not belong to the influencing parameter;

[0136] It should be noted that the value range of the correlation coefficient is [-1,1]. The correlation coefficient is selected by t, and the value range of t is [0,1]. The larger t is, the fewer the number of selected parameters is, and the greater the correlation between the selected parameters and fuel consumption is; the smaller t is, the more the number of selected parameters is, and the smaller the correlation between the selected parameters and fuel consumption is. When making a specific selection, the value is limited according to the number of obtained ship parameters and climate parameters.

[0137] Step S232: Obtain the correlation coefficient XG(qz) between the climate parameter and the ship fuel consumption, and make a judgment and analysis on XG(qz) in combination with the correlation coefficient threshold t:

[0138] If |XG(qz)|≥t, the climate parameter is an influencing parameter and the climate parameter is extracted;

[0139] If |XG(qz)|<t, then the climate parameter is not an influencing parameter;

[0140] Step S3: According to the influencing parameters, the forces of the influencing parameters on the ship are obtained, the forces are integrated to obtain the total forces, and regression analysis is performed based on the total forces combined with the fuel consumption information to construct a fuel consumption prediction model.

[0141] Step S31: Please refer to Figure 2 , obtain the force F(i, cy) and F(i, qz) of the influencing parameters on the ship; obtain the angle θ between the force and the ship's travel direction cy ,θ qz , obtain the friction coefficient xs of the hull; integrate the forces to obtain the total force zF i ;

[0142] zF i =F(i,cy)cosθ cy +xs×F(i,cy)sinθ cy +F(i,qz)cosθ qz +xs×F(i,qz)sinθ qz ;

[0143] It should be noted that the force of the influencing parameters on the ship is decomposed through the angle, so that it is mainly composed of horizontal and vertical forces. The vertical force is converted through the friction with the water surface and unified into the horizontal force.

[0144] This formula normalizes various external factors in the ship's navigation. All factors are calculated uniformly according to the force, and decomposed according to the direction and angle of the force. The sum of the forces affected by all factors is obtained, and combined with the ship's fuel consumption information, the relationship between the ship's fuel consumption change and the total effect is reflected, which is convenient for model construction.

[0145] For example, for the first ship, it is mainly affected by its load, and F(i, cy)=10 is obtained according to its load. 9 , during the voyage, the main wind is 45° in the opposite direction, and the wind force is level 6, then F(i, qz) = 10 6 , obtain the friction coefficient xs = 0.02, then the total force is:

[0146]

[0147] Step S32: Please refer to Figure 3 ;According to the fuel consumption information, obtain the fuel consumption xh of the ship i , combined with the total force, construct the total force zF i Fuel consumption xh i The data set is processed by least squares to obtain the relevant parameters of the fuel consumption prediction model;

[0148] The specific process is as follows:

[0149] Step S321: Obtain data set, based on the total force zF i Fuel consumption xh iPerform calculations to obtain the first parameter k of the fuel consumption prediction model;

[0150] The specific calculation process of the first parameter k is as follows:

[0151]

[0152] It should be noted that: i Refers to the total force acting on the ith ship, xh i Refers to the fuel consumption generated by the i-th ship;

[0153] Step S322: According to the total force zF i Fuel consumption xh i , combined with the first parameter k, the second parameter v of the fuel consumption prediction model is calculated;

[0154] The specific calculation process of the second parameter v is as follows:

[0155]

[0156] Step S323: Obtain the first parameter k and the second parameter v, and construct a fuel consumption prediction model. The fuel consumption prediction model is as follows:

[0157] xh=k×zF+v;

[0158] Step S4: Perform navigation prediction and real-time prediction of the fuel consumption of the ship according to the fuel consumption prediction model, prepare the fuel volume of the ship according to the prediction results, and issue an alarm for abnormal fuel consumption generated during navigation;

[0159] Step S41: Before the ship is ready to set sail, the navigation route is obtained through the navigation system; based on the weather forecast data provided by the Meteorological Bureau, including wind speed, wind direction, wave height and possible severe weather warnings, as well as the relevant sea conditions information released by the Maritime Administration, the climate conditions are obtained, and the natural forces acting on the hull are estimated using a mechanical model to obtain the total force zF1 acting on the hull during navigation; combined with the fuel consumption prediction model, the total force zF1 is brought into the model for calculation to obtain the fuel consumption xh1 required for navigation; based on xh1, sufficient fuel is prepared for the ship to prevent the navigation plan or safety from being affected by insufficient fuel.

[0160] Step S42: When the ship is in the sailing state, the fuel consumption is monitored in real time; high-precision sensors and measuring instruments are used to continuously monitor and record the various forces actually applied to the hull during the sailing process, including wind force, water flow impact force, and turbulence caused by waves, and the force zF2 applied during the sailing process is obtained. The predicted fuel consumption xh2 is obtained by combining the fuel consumption prediction model, and the actual fuel consumption xhs of the ship at the current moment is obtained; the difference between the predicted fuel consumption and the actual fuel consumption is compared. If xhs>xh2, the fuel consumption warning mechanism is triggered, and the crew needs to inspect the hull and investigate the cause of the abnormal fuel consumption.

[0161] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technical personnel in this field according to actual conditions. For example, if there are weight coefficients and proportional coefficients, their set sizes are to quantify each parameter to obtain a specific value, which is convenient for subsequent comparison. Regarding the size of the weight coefficient and the proportional coefficient, it is sufficient as long as it does not affect the proportional relationship between the parameter and the quantized value.

[0162] Finally, it should be noted that the above-described embodiments are only specific implementations of the present invention, which are used to illustrate the technical solutions of the present invention, rather than to limit them. The protection scope of the present invention is not limited thereto. Although the present invention is described in detail with reference to the above-described embodiments, ordinary technicians in the field should understand that any technician familiar with the technical field can still modify the technical solutions recorded in the above-described embodiments within the technical scope disclosed by the present invention, or can easily think of changes, or make equivalent replacements for some of the technical features therein; and these modifications, changes or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention shall be based on the protection scope of the claims.

Claims

1. A method for predicting fuel consumption of ocean-going ships based on big data, characterized in that: The prediction method comprises: Step S1: Obtain ship data and climate data, and obtain fuel consumption information; Step S2: Obtain ship parameters and climate parameters according to ship data and climate data, process the ship parameters and climate parameters, extract abnormal values ​​in the ship parameters and climate parameters, calculate the correlation coefficient between the ship parameters and climate parameters in combination with fuel consumption information, analyze and judge the correlation coefficient, and obtain the influencing parameters; Step S3: Obtain the force of the influencing parameters on the ship, perform force analysis on the force on the ship, obtain the total force, perform regression analysis based on the total force combined with the fuel consumption information, and build a fuel consumption prediction model; Step S4: Perform navigation prediction and real-time prediction of the fuel consumption of the ship according to the fuel consumption prediction model, and issue an alarm for abnormal fuel consumption generated during the navigation process.

2. The method for predicting fuel consumption of ocean-going ships based on big data according to claim 1 is characterized in that: The specific steps of step S2 are as follows: Step S21: Processing the ship data and climate data to obtain ship parameters and climate parameters, performing numerical analysis on the ship parameters and climate parameters to obtain the distribution of the ship parameters and climate parameters, and extracting abnormal values ​​in the ship parameters and climate parameters according to the distribution; Step S22: according to the processed data, the mean of the ship parameters and the climate parameters is obtained, and the correlation coefficient between the ship parameters, the climate parameters and the fuel consumption is calculated in combination with the fuel consumption information; Step S23: Analyze and judge the correlation coefficient, extract the ship parameters and climate parameters with large correlation coefficients, and obtain the influencing parameters.

3. The method for predicting fuel consumption of ocean-going ships based on big data according to claim 2 is characterized in that: The specific steps of step S21 are as follows: Step S211: Obtain ship parameter type y, obtain data of x ships according to ship data, and obtain ship parameter cb(i, cy), 0<cy≤y, cy refers to the specific ship parameter type; obtain the climate type z to which the ship is subject, obtain the climate impact on x ships according to the climate data, and obtain climate parameter qh(i, qz), 0<qz≤z, qz refers to the specific ship parameter type, and i refers to the i-th ship; Step S212: Perform numerical analysis on the ship parameter cb(i, cy) and the climate parameter qh(i, qz), and determine whether the parameters conform to the normal distribution according to the analysis results; Step S213: extracting abnormal values ​​according to the judgment result; If the parameters conform to the normal distribution, outliers are extracted according to the 3σ principle; If the parameters do not conform to the normal distribution, outliers are extracted according to the isolation forest method.

4. The method for predicting fuel consumption of ocean-going ships based on big data according to claim 3 is characterized in that: The specific steps of step S212 are as follows: Step S2121: Calculate the mean of the ship parameters to obtain the mean μ(cy) of the ship parameters; Step S2122: According to the ship parameters, combined with the mean of the ship parameters, the standard deviation of the ship parameters is obtained to obtain the ship parameter standard deviation σ(cy): Step S2123: According to the mean value μ(cy) of the ship parameters, the standard deviation σ(cy) of the ship parameters, and in combination with the ship parameters, the kurtosis and skewness of the ship parameters are obtained to obtain the ship parameter kurtosis FD(cy) and the ship parameter skewness PD(cy): Step S2124: Analyze and judge the ship parameters according to the ship parameter kurtosis FD(cy) and the ship parameter skewness PD(cy): If -2≤FD(cy)≤2, and -0.5≤PD(cy)≤0.5, the ship parameters conform to the normal distribution; If FD(cy)>2 or FD(cy)<-2 or PD(cy)>0.5 or PD(cy)<-0.5, the ship parameters do not conform to the normal distribution.

5. The method for predicting fuel consumption of ocean-going ships based on big data according to claim 4 is characterized in that: The step S212 further includes: Step S2125: Calculate the mean of the climate parameter to obtain the climate parameter mean μ(qz); Step S2126: according to the climate parameter and in combination with the mean value of the climate parameter, the standard deviation of the climate parameter is obtained to obtain the climate parameter standard deviation σ(qz); Step S2127: according to the climate parameter mean μ(qz), the climate parameter standard deviation σ(qz), and the climate parameter, the kurtosis and skewness of the climate parameter are calculated in combination with the climate parameter to obtain the climate parameter kurtosis FD(qz) and the climate parameter skewness PD(qz); By obtaining the kurtosis and skewness of the climate parameters, the data distribution of the climate parameters is judged, and the parameters that conform to the normal distribution are extracted through the kurtosis and skewness data; Step S2128: Analyze and judge the climate parameters according to the climate parameter kurtosis FD(qz) and the climate parameter skewness PD(qz): If -2≤FD(qz)≤2, and -0.5≤PD(qz)≤0.5, the climate parameter conforms to the normal distribution; If FD(qz)>2 or FD(qz)<-2 or PD(qz)>0.5 or PD(qz)<-0.5, the climate parameter does not conform to the normal distribution.

6. The method for predicting fuel consumption of ocean-going ships based on big data according to claim 2 is characterized in that: The specific steps of step S22 are as follows: Step S221: According to the fuel consumption information, the fuel consumption of the ship YH(i, cy, qz) is obtained, and the average fuel consumption of the ship under different climates and the average fuel consumption under different ship data are obtained to obtain the average fuel consumption of the ship CJZ and the average fuel consumption of the climate QJZ; The calculation process of the ship fuel consumption mean CJZ is as follows: YH(i, cy, qz) refers to the fuel consumption of the i-th ship under the cy-th ship parameters and under the qz-th climate influence, x refers to the number of ships, and z refers to the climate type; The calculation process of climate fuel consumption mean QJZ is as follows: Among them, y refers to the type of ship parameters; Step S222: Obtain ship parameters cb(i, cy) and climate parameters qh(i, qz), and obtain the mean values ​​of ship parameters J(cy) and climate parameters J(qz); Step S223: according to the ship parameter cb(i, cy), the mean value of the ship parameter J(cy), the fuel consumption of the ship YH(i, cy, qz), and the mean value of the fuel consumption of the ship CJZ, the correlation between the ship parameter and the fuel consumption of the ship is obtained to obtain the correlation coefficient XG(cy); Step S224: according to the climate parameter qh(i, qz), the climate parameter mean J(qz), the fuel consumption of the ship YH(i, cy, qz), and the climate fuel consumption mean CJZ, the correlation between the climate parameter and the ship fuel consumption is obtained to obtain the correlation coefficient XG(qz); 7. The method for predicting fuel consumption of ocean-going ships based on big data according to claim 2 is characterized in that: The subsequent steps of step S23 are as follows: Step S231: Obtain the correlation coefficient threshold t, 0<t≤1, obtain the correlation coefficient XG(cy) between the ship parameters and the ship fuel consumption, and make a judgment and analysis on XG(cy) in combination with the correlation coefficient threshold t: If |XG(cy)|≥t, the ship parameter is an influencing parameter and the ship parameter is extracted; If |XG(cy)|<t, then the ship parameter does not belong to the influencing parameter; Step S232: Obtain the correlation coefficient XG(qz) between the climate parameter and the ship fuel consumption, and make a judgment and analysis on XG(qz) in combination with the correlation coefficient threshold t: If |XG(qz)|≥t, the climate parameter is an influencing parameter and the climate parameter is extracted; If |XG(qz)|<t, then the climate parameter is not an influencing parameter.

8. The method for predicting fuel consumption of ocean-going ships based on big data according to claim 1 is characterized in that: The specific steps of step S3 are as follows: Step S31: Obtain the forces F(i, cy) and F(i, qz) of the influencing parameters on the ship; obtain the angle θ between the force and the ship's travel direction cy ,θ qz , obtain the friction coefficient xs of the hull; Integrate the forces to get the total force zF i ; zF i =F(i,cy)cosθ cy +xs×F(i,cy)sinθ cy +F(i,qz)cosθ qz +xs×F(i,qz)sinθ qz ; Step S32: Obtain the fuel consumption xh of the ship according to the fuel consumption information i , combined with the total force, construct the total force zF i Fuel consumption xh i The data set is processed by least squares to obtain the first parameter k and the second parameter v of the fuel consumption prediction model; According to the first parameter k and the second parameter v, a fuel consumption prediction model is constructed to obtain a fuel consumption prediction model. The fuel consumption prediction model is as follows: xh=k×zF+v xh is the fuel consumption variable of the fuel consumption prediction model, and zF is the total force variable of the fuel consumption prediction model.

Images