Bayesian optimization-based ship CII rating optimization method and system

Abstract

The invention provides a ship CII rating optimization method and system based on Bayesian optimization, and the method comprises a multi-source data fusion step, a ship oil consumption calculation model construction step, a speed optimization step, a ship CII rating calculation step, a CII rating optimization model construction step, and an optimization result output step. Multi-source data are fused to construct historical operation data of a ship, speed optimization is carried out to carry out CII rating on an operating ship, Bayesian optimization is adopted to establish a Bayesian optimization model based on ship navigation performance, economic benefit maximization and CII rating as constraints, and balance between ship operation maximization and influence on the environment is realized. And meanwhile, the requirements of the international maritime organization on CII are met, and navigation reference is provided for providing navigational speed and load factor indexes for green low-carbon navigation of ships.

Description

Technical Field

[0001] The present invention relates to the field of ship engineering technology, and in particular to a ship CII rating optimization method and system based on Bayesian optimization. Background Art

[0002] Shipping, as an important mode of transport for people and goods, accounts for more than 80% of global trade, and also generates a large amount of greenhouse gas (GHG) emissions, which puts serious pressure on the global marine and atmospheric environment. According to the latest "Fourth IMO 2020 Greenhouse Gas Study" published by the International Maritime Organization (IMO), total shipping CO2 showed an increase of 9.77%, from 962 million tons in 2012 to 1.056 billion tons in 2018. Last year, the share of shipping emissions in global anthropogenic emissions reached 2.89%. The importance of timely and effective measures to reduce greenhouse gas emissions from transportation cannot be overemphasized.

[0003] The International Maritime Organization is responsible for ship safety and the prevention of marine and atmospheric pollution from ships. Its Marine Environment Protection Committee (MEPC) has been promoting the reduction of shipping emissions by setting strategic goals and a series of regulatory measures. The latter includes the Carbon Intensity Index (CII) in Operation, which has attracted widespread attention in the shipping industry since its introduction. The CII, adopted by the MEPC in June 2021 and effective from January 1, 2023, specifies the calculation formula for quantifying carbon intensity levels and establishing a rating system for operating ships.

[0004] The International Maritime Organization points out that a ship running on low or zero carbon fuels can get a higher rating than one running on fossil fuels, but there are many other measures that can be taken to improve a ship's rating, involving a variety of technical and operational solutions. In terms of technical solutions, typical measures include the introduction of engine power limitation (Engine Power Limitation, referred to as EPL), hull cleaning, energy saving devices (Energy Saving Devices, ESDs) and the application of waste heat recovery (WHR) systems.

[0005] There are potential disadvantages in introducing these measures, among which, EPL: may affect the speed and flexibility of the ship, and reduce the ability to respond in emergencies, and may increase operating time and fuel costs in the long run; hull cleaning: requires regular maintenance, increasing operational complexity and costs; ESDs: installation and maintenance costs are high, which may affect the space utilization and operational performance of the ship; WHR system: large initial investment, complex system, maintenance and operation require professional skills, and will also affect the overall space and weight distribution of the ship. In other words, technical measures such as ESDs installation and hull cleaning take time to actually implement, and most importantly, require upfront investment. In contrast, speed optimization is a more direct way without additional investment. However, meeting the CII rating requirements to ensure the continuous operation of the ship is only a basic condition. How to maximize economic benefits while meeting the IMO's carbon reduction standards is the most concerned issue for ship owners, charter companies and other stakeholders. Summary of the invention

[0006] In order to solve the problems that existing measures to improve the rating of ships have the effects on the flexibility of ships, high installation and maintenance costs, large initial investment and complex systems, the present invention provides a ship CII rating optimization method based on Bayesian optimization, which integrates multi-source data to construct historical ship operation data and speed optimization to perform CII rating on operating ships, and adopts Bayesian optimization to establish a Bayesian optimization model based on ship navigation performance, economic benefit maximization and CII rating as constraints, so as to balance between ship operation maximization and environmental impact, maximize operational benefits, meet the requirements of the International Maritime Organization for CII, and provide speed and full load factor indicators for green and low-carbon navigation of ships. Provide navigation reference. The present invention also relates to a ship CII rating optimization system based on Bayesian optimization.

[0007] The technical solution of the present invention is as follows:

[0008] A ship CII rating optimization method based on Bayesian optimization, characterized by comprising the following steps:

[0009] The multi-source data fusion step is to fuse and analyze the ship AIS data and MRV message data from several data sources to form the basic data of the ship's historical operation;

[0010] A ship fuel consumption calculation model construction step, based on the ship's historical operation basic data, combined with the cubic relationship between the ship's fuel consumption and the ship's rotation speed from the MRV message data, to establish a ship fuel consumption calculation model, and calculate the ship's fuel consumption and the ship's carbon emissions according to the ship fuel consumption calculation model;

[0011] a speed optimization step, based on the historical operation basic data of the ship, dividing the historical sailing speed of the ship from the ship AIS data into several speed levels, and calculating the average speed of different speed levels, combining the proportion of the sailing time of the ship at different speed levels, the proportion of the distance of the ship in the empty and full-load states, and the average speed of different speed levels to calculate the sailing distance of the ship in the empty and full-load states, and converting the calculated ship fuel consumption into the fuel consumption of the speed-optimized ship in the empty and full-load states;

[0012] The ship CII rating calculation step is to calculate the ship CII rating for a given year based on the historical operation basic data of the ship and the calculated ship carbon emissions;

[0013] The steps of constructing the CII rating optimization model are as follows: based on the calculated fuel consumption of the ship under empty and full load conditions with optimized speed, the sailing distance of the ship under empty and full load conditions, and the CII rating of the ship in a given year, a CII rating optimization model is established using the Bayesian optimization algorithm with the ship's CII rating as a constraint and the maximization of the sailing distance of the ship under full load conditions as the optimization goal;

[0014] The optimization result output step optimizes the CII rating of the ship according to the constructed CII rating optimization model, and outputs the optimization results of the average speed of different speed levels and the optimization results of the sailing distance under empty and full load conditions.

[0015] Preferably, in the step of constructing the ship fuel consumption calculation model, based on the cubic relationship between the ship fuel consumption and the ship speed derived from the MRV message data in the ship's historical operating basic data and the empty and full load states determined by the draft data, a ship fuel consumption calculation model is established and the ship's fuel consumption in the full load state and the ship's fuel consumption in the empty load state as well as the corresponding ship carbon emissions are calculated.

[0016] Preferably, in the speed optimization step, based on the historical operation basic data of the ship, the historical sailing speed of the ship derived from the ship AIS data is divided into three levels: low, medium and high, and the average speeds of the ships at the low, medium and high levels are calculated respectively; the sailing distance of the ship in the empty and full-load states is calculated in combination with the respective sailing time proportions of the ships at the low, medium and high levels, the total sailing time throughout the year, the proportion of the distance of the ships in the empty and full-load states, and the average speeds of the ships at the low, medium and high levels; and according to the respective sailing time proportions of the ships at the low, medium and high levels, the total sailing time throughout the year, the proportion of the distance of the ships in the empty and full-load states, and the average speeds of the ships at the low, medium and high levels, the calculated fuel consumption of the ship in the full-load state and the fuel consumption of the ship in the empty-load state are respectively converted into the fuel consumption of the ships in the empty and full-load states with optimized speeds of each level.

[0017] Preferably, in the ship CII rating calculation step, the product of the ship carrying capacity and the sailing distance in a given year is taken as the total transport work undertaken, and the ship CII value for the given year is calculated by the ratio of the calculated ship carbon emissions to the total transport work undertaken; and the curve of the median of the operational carbon intensity performance of the ship in a certain historical year is taken as the CII baseline reference value, and the required ship CII value for the given year is calculated according to the requirements of the International Maritime Organization on carbon emissions and in combination with the CII baseline reference value and the CII rating reduction factor; and then the ratio between the ship CII value for the given year and the required ship CII value is calculated and according to the CII rating calculation rules, the CII rating of the ship in the given year is obtained.

[0018] Preferably, in the step of constructing the CII rating optimization model, a Bayesian optimization algorithm is used, and the average speeds of the three grades of ships, low, medium and high, and the respective sailing time proportions of the three grades of ships are taken as decision variables, and the current annual ship CII value is less than or equal to the higher boundary value in the CII rating calculation rule, the average speed of the low-level ship is less than the 33% quantile of the historical daily speed, the average speed of the medium-level ship is between the 33% quantile and the 66% quantile of the historical daily speed, and the average speed of the high-level ship is between the 66% quantile of the historical daily speed and the maximum value of the historical daily speed, and a CII rating optimization model is established with the ship CII rating as a constraint condition and maximizing the sailing distance of the ship under full load as the optimization goal.

[0019] Preferably, in the optimization result output step, the ship's target CII rating is set based on the CII rating calculation rules, the ship's target CII rating is input into the established CII rating optimization model and a set of feasible solutions are calculated and solved according to the preset initial value and number of iterations to complete the optimization of the ship's CII rating, and the optimization results of the ship's average speed and corresponding sailing time at different speed levels in the current year and the optimization results of the sailing distance under empty and full-load states are output.

[0020] A ship CII rating optimization system based on Bayesian optimization, characterized by comprising a multi-source data fusion module, a ship fuel consumption calculation model construction module, a speed optimization module, a ship CII rating calculation module, a CII rating optimization model construction module and an optimization result output module connected in sequence,

[0021] The multi-source data fusion module performs multi-source data fusion and analysis on the ship AIS data and MRV message data from several data sources to form basic data of historical ship operations;

[0022] The ship fuel consumption calculation model building module establishes a ship fuel consumption calculation model based on the ship historical operation basic data and the cubic relationship between the ship fuel consumption and the ship rotation speed derived from the MRV message data, and calculates the ship fuel consumption and the ship carbon emissions according to the ship fuel consumption calculation model;

[0023] The speed optimization module, based on the historical operation basic data of the ship, divides the historical sailing speed of the ship from the ship AIS data into several speed levels, calculates the average speed of different speed levels, calculates the sailing distance of the ship in the empty and full-load state by combining the proportion of the sailing time of the ship at different speed levels, the proportion of the distance of the ship in the empty and full-load state, and the average speed of different speed levels, and converts the calculated fuel consumption of the ship into the fuel consumption of the ship in the empty and full-load state with speed optimization;

[0024] The ship CII rating calculation module calculates the ship CII rating for a given year based on the ship's historical operating basic data combined with the calculated ship carbon emissions;

[0025] The CII rating optimization model building module uses a Bayesian optimization algorithm to establish a CII rating optimization model with the ship's CII rating as a constraint and maximizing the ship's sailing distance under a fully loaded state as an optimization goal, based on the calculated fuel consumption of the ship under the speed-optimized empty and full-load states, the sailing distance of the ship under the empty and full-load states, and the CII rating of the ship in a given year;

[0026] The optimization result output module optimizes the CII rating of the ship according to the constructed CII rating optimization model, and outputs the optimization results of the average speed of different speed levels and the optimization results of the sailing distance under empty and full load states.

[0027] Preferably, in the ship fuel consumption calculation model construction module, based on the cubic relationship between the ship fuel consumption and the ship rotation speed derived from the MRV message data in the ship's historical operation basic data and the empty and full load states determined by the draft data derived from the ship's AIS data, a ship fuel consumption calculation model is established to calculate the ship's fuel consumption in the full load state and the ship's fuel consumption in the empty load state as well as the corresponding ship carbon emissions.

[0028] Preferably, in the speed optimization module, based on the historical operation basic data of the ship, the historical sailing speed of the ship derived from the ship AIS data is divided into three levels: low, medium and high, and the average speeds of the ships at the low, medium and high levels are calculated respectively; the sailing distance of the ship in the empty and full-load states is calculated by combining the respective sailing time proportions of the ships at the low, medium and high levels, the total sailing time throughout the year, the proportion of the distance of the ships in the empty and full-load states, and the average speeds of the ships at the low, medium and high levels; and according to the respective sailing time proportions of the ships at the low, medium and high levels, the total sailing time throughout the year, the proportion of the distance of the ships in the empty and full-load states, and the average speeds of the ships at the low, medium and high levels, the calculated fuel consumption of the ship in the full-load state and the fuel consumption of the ship in the empty-load state are respectively converted into the fuel consumption of the ship in the empty and full-load states with optimized speed of each level.

[0029] Preferably, in the ship CII rating calculation module, the product of the ship carrying capacity and the sailing distance in a given year is taken as the total transport work undertaken, and the ship CII value in the given year is calculated by the ratio of the calculated ship carbon emissions to the total transport work undertaken; and the curve of the median of the operational carbon intensity performance of the ship in a certain historical year is taken as the CII baseline reference value, and the required ship CII value in the given year is calculated according to the requirements of the International Maritime Organization on carbon emissions and in combination with the CII baseline reference value and the CII rating reduction factor; and then the ratio between the ship CII value in the given year and the required ship CII value is calculated and according to the CII rating calculation rules, the CII rating of the ship in the given year is obtained.

[0030] The beneficial effects of the present invention are:

[0031] The present invention provides a ship CII rating optimization method based on Bayesian optimization, collects ship AIS data, MRV message data and other data from several data sources for multi-source data fusion and analysis, forms ship historical operation basic data, and provides a ship CII calculation and rating method based on historical operation data and Bayesian optimization, calculates ship fuel consumption and ship carbon emissions based on the ship historical operation basic data, and then divides the ship's historical navigation speed into several speed levels based on the ship's historical operation basic data, and calculates the average speed of different speed levels, and calculates the navigation distance of the ship in the empty and full load state by combining the proportion of the ship's navigation time at different speed levels, the proportion of the ship's empty and full load state distance and the average speed of different speed levels, and converts the calculated ship fuel consumption into the fuel consumption of the ship in the empty and full load state of the speed-optimized ship, and the speed optimization step can be understood as redistributing the proportion of different speed ranges and adjusting the corresponding average values ​​to establish a speed optimization model for operating ships, so that it can provide guidance for operational decisions under increasingly stringent CII regulations. Speed ​​optimization is a direct method to improve energy efficiency. By adjusting the speed, fuel consumption and carbon emissions can be significantly reduced. Appropriate speed adjustment can optimize fuel use without increasing investment in other equipment and technology, thereby improving the CII rating. According to the calculated fuel consumption of the ship under empty and full load conditions, the sailing distance of the ship under empty and full load conditions, and the CII rating of the ship in a given year, a Bayesian optimization algorithm is used to establish a CII rating optimization model (also known as a Bayesian optimization model) based on ship navigation performance, economic benefit maximization, and CII rating as constraints, with maximizing the sailing distance of the ship under full load as the optimization goal. The model strikes a balance between maximizing the economic benefits of ship operation and the impact on the environment, and adopts a method based on Bayesian optimization of CII rating to maximize economic benefits while meeting the requirements of the International Maritime Organization for CII. After optimizing the CII rating of the ship according to the constructed CII rating optimization model, the optimization results of the average speed of different speed levels and the optimization results of the sailing distance under empty and full load conditions are output. The present invention can also be called a ship CII rating optimization algorithm based on historical operating basic data and Bayesian optimization. It performs CII rating on operating ships based on historical ship operating data. It proposes a method for maximizing the economic benefits of ships while meeting CII requirements, and involves technologies related to ship greenhouse gas emission reduction and energy efficiency. Through the Bayesian optimization algorithm, a speed optimization model for operating ships is established to maximize economic benefits while meeting CII standards. It can be applied to ship engineering, carbon emission optimization, artificial neural networks, CII rating, etc., and can improve the energy efficiency of ships and reduce carbon emissions to meet environmental regulations and reduce operating costs. The ship CII rating based on Bayesian optimization provides speed and full load factor indicator recommendations for green and low-carbon navigation of ships.

[0032] The present invention also relates to a ship CII rating optimization system based on Bayesian optimization, which corresponds to the above-mentioned ship CII rating optimization method based on Bayesian optimization, and can be understood as a system for realizing the above-mentioned ship CII rating optimization method based on Bayesian optimization, including a multi-source data fusion module, a ship fuel consumption calculation model construction module, a speed optimization module, a ship CII rating calculation module, a CII rating optimization model construction module and an optimization result output module connected in sequence, each module works in coordination with each other, and integrates several types of data such as ship AIS data and MRV message data to form ship historical operation data, so the present invention can be understood as It provides a ship CII calculation and rating system based on historical operating data, integrates multi-source data to construct ship historical operating data and optimizes speed to conduct CII rating for operating ships, and uses Bayesian optimization to establish a Bayesian optimization model based on ship navigation performance, economic benefit maximization, and CII rating as constraints. It strikes a balance between ship operation economic maximization and environmental impact, and uses a system based on Bayesian optimization CII rating to maximize economic benefits while meeting the International Maritime Organization's requirements for CII. The ship CII rating based on Bayesian optimization provides a navigation reference for the navigation speed of the ship under empty and fully loaded conditions. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1 It is a flow chart of the ship CII rating optimization method based on Bayesian optimization of the present invention.

[0034] Figure 2 It is the result of ship CII rating optimization based on Bayesian optimization of the present invention.

[0035] Figure 3 It is a structural block diagram of the ship CII rating optimization system based on Bayesian optimization of the present invention. DETAILED DESCRIPTION

[0036] The present invention will be described below in conjunction with the accompanying drawings.

[0037] The present invention relates to a ship CII rating optimization method based on Bayesian optimization, and the flow chart of the method is as follows: Figure 1 As shown, the following steps are included:

[0038] S1, multi-source data fusion step, multi-source data fusion and analysis of ship AIS data and MRV message data from multiple data sources to form basic data of ship historical operation.

[0039] This step is specifically multi-source data fusion. By selecting several types of multi-source data such as ship AIS data, MRV message data, meteorological data, etc. for fusion and analysis, the basic data of ship historical operation is formed as a data set for subsequent optimization of ship CII rating. Therefore, the present invention can also be understood as a ship CII calculation and rating method based on historical operation data.

[0040] The ship AIS data, MRV message data and meteorological data in this dataset - ship historical operation basic data are important sources of information for optimizing ship navigation and improving energy efficiency. In addition to basic ship information, this dataset also mainly includes the daily sailing time, sailing distance and loading status of each ship.

[0041] Among them, ship AIS data provides information about the real-time position and navigation status of the ship, such as: ship identification: basic ship information such as the ship's MMSI (Maritime Mobile Service Identification Code), ship name, call sign, IMO (International Maritime Organization) number; location data: latitude and longitude coordinates (GPS location), navigation track, current location; speed and heading: the ship's current speed (in knots) and heading (in degrees); ship type: the category and function of the ship, such as tanker, container ship, cargo ship, etc.; navigation status: such as sailing, moored, anchored, etc.; destination port and estimated arrival time: the ship's destination port and estimated arrival time.

[0042] MRV message data, i.e. the message data of the monitoring, reporting and verification (MRV) system, mainly involves the energy consumption and emissions of ships, such as: fuel consumption and related operation data: the fuel consumption and ship rotation speed of the ship during navigation, usually classified by fuel type (such as diesel, heavy oil, etc.); emission data: including emissions of pollutants such as carbon dioxide (CO2), nitrogen oxides (NOx), sulfur oxides (SOx); performance energy efficiency indicators: such as draft data, fuel consumption per unit of cargo transportation (such as ton-mile fuel consumption) or energy efficiency ratio; sailing time: the sailing time and berthing time of the ship in different sections; sailing distance: the total sailing distance of the ship during the reporting period; reporting period: the time period for data recording, such as daily, weekly or monthly.

[0043] Preferably, meteorological data can also be integrated, which provides information on the environmental conditions when the ship is sailing, such as: wind speed and direction: wind speed (usually in knots or meters per second) and wind direction (in degrees); wave height: average height of waves and maximum wave height; air temperature and humidity: air temperature and humidity level; air pressure: measurement of atmospheric pressure; precipitation: amount of precipitation (such as rainfall); visibility: line of sight or visibility conditions, which affect navigation safety; weather forecast: including future weather trends and warning information, such as storms, typhoons, etc.

[0044] S2, a step of constructing a ship fuel consumption calculation model, based on the historical operation basic data of the ship and the cubic relationship between the ship fuel consumption and the ship rotation speed derived from the MRV message data, a ship fuel consumption calculation model is established, and the ship fuel consumption and the ship carbon emissions are calculated according to the ship fuel consumption calculation model.

[0045] Furthermore, based on the classic cubic relationship between the ship fuel consumption and the ship rotation speed derived from the MRV message data in the ship's historical operating basic data and the empty and full load states determined by the draft data, a ship fuel consumption calculation model can be established to calculate the ship's fuel consumption in the fully loaded state and the ship's fuel consumption in the empty state and the corresponding ship carbon emissions respectively.

[0046] Among them, the product of the ship's fuel consumption and the carbon emission factor is the ship's carbon emissions. The carbon emission factor is a carbon emission factor specific to the fuel type, usually in tons of CO2 per ton of fuel (tCO2 / t fuel). Different fuels have different emission factors. For example: heavy fuel oil (HFO): about 3.114tCO2 / t fuel, diesel (MDO / MGO): about 3.206tCO2 / t fuel, natural gas (LNG): about 2.75tCO2 / t fuel.

[0047] S3, a speed optimization step, based on the historical operation basic data of the ship, dividing the historical sailing speed of the ship from the ship AIS data into several speed grades, and calculating the average speed of different speed grades, combining the proportion of the ship's sailing time at different speed grades, the proportion of the ship's empty and full-load state distance and the average speed of different speed grades to calculate the sailing distance of the ship in the empty and full-load state, and converting the calculated ship fuel consumption into the speed-optimized ship's fuel consumption in the empty and full-load state.

[0048] Specifically, the historical sailing speeds of ships derived from the ship AIS data can be divided into three levels: low, medium and high, and the average speeds of the ships at the low, medium and high levels are calculated respectively. The sailing distance of the ship in the empty and full-load states is calculated by combining the respective sailing time proportions of the ships at the low, medium and high levels, the total sailing time throughout the year, the proportion of the distance of the ships in the empty and full-load states, and the average speeds of the ships at the low, medium and high levels; and according to the respective sailing time proportions of the ships at the low, medium and high levels, the total sailing time throughout the year, the proportion of the distance of the ships in the empty and full-load states, and the average speeds of the ships at the low, medium and high levels, the calculated fuel consumption of the ship in the full-load state and the fuel consumption of the ship in the empty-load state are respectively converted into the fuel consumption of the ship in the empty and full-load states with optimized speeds of each level.

[0049] The speed optimization step can also be understood as being used to establish a speed optimization model. The historical sailing speed of the ship is divided into three levels: low, medium and high, and the corresponding sailing distance is calculated in combination with the empty and full load states.

[0050] (1) The sailing distance D of the ship when empty or fully loaded i for:

[0051]

[0052] Where i=1, 2, 3, β1, β2, β3 correspond to the proportion of low, medium and high speed sailing time of the ship respectively; p represents the ratio of full load state to total sailing distance, V1, V2, V3 correspond to the average speed of low, medium and high levels of the ship respectively, and T represents the total sailing time throughout the year.

[0053] (2) Fuel consumption FC of the ship under empty and full load conditions after speed optimization i for:

[0054]

[0055] Among them, f e (V i ) and f f (V i ) represent the fuel consumption of the ship under no-load and full-load conditions calculated in the ship fuel consumption calculation model construction step.

[0056] S4, a ship CII rating calculation step, based on the historical operation basic data of the ship and the calculated ship carbon emissions, calculates the ship CII rating for a given year.

[0057] In the ship CII rating calculation step, the product of the ship carrying capacity and the sailing distance in a given year is taken as the total transport work undertaken, and the ship CII value for the given year is calculated by the ratio of the calculated ship carbon emissions to the total transport work undertaken; and the curve of the median of the operational carbon intensity performance of the ship in a certain historical year is taken as the CII baseline reference value, and the ship CII value required for the given year is calculated according to the requirements of the International Maritime Organization on carbon emissions and in combination with the CII baseline reference value and the CII rating reduction factor; and then the ratio between the ship CII value for the given year and the required ship CII value is calculated and according to the CII rating calculation rules, the ship CII rating for the given year is obtained.

[0058] (1) Calculate the CII value of a ship in a given year based on the ship's historical operating data.

[0059]

[0060] Where M refers to all fuel oil consumed on board a ship in a given year (given calendar year) j The total mass of the total CO2 emissions generated (g), C Fjrepresents the carbon emission factor coefficient, usually 3.114; W is the total transport work undertaken, which is determined by the ship's carrying capacity Capacity and the sailing distance D in a given calendar year. t The product of is obtained.

[0061] (2) CII Baseline Calculation. The Operational Carbon Intensity Index (CII) baseline is defined as a curve representing the median operational carbon intensity performance achieved by a specified group of ships in 2019, as a function of carrying capacity.

[0062] CII ref =a*Capacity -c

[0063] The values ​​of parameters a and c depend on different ship types, see IMO CII Guidelines G2 for details.

[0064] (3) Calculate the CII value required by the IMO CII Guidelines.

[0065] Required CII=(1-Z)*CII ref

[0066] Among them, the corresponding CII rating reduction factor z value for each year from 2023 to 2030 can be found in the IMO CII Guide G3. As shown in Table 1, the corresponding reduction factor Z values ​​for 2023-2026 are 5%, 7%, 9% and 11% respectively.

[0067] Table 1

[0068]

[0069] (4) CII rating calculation. The CII rating can be calculated by the ratio between the Attained CII and the Required CII. The CII rating is divided into five levels, A, E, and is defined by four boundaries. IMO requires that ships must reach C or above. The CII rating calculation rules are as follows:

[0070]

[0071] Among them, Superior Boundary is the excellent boundary, Lower Boundary is the lower boundary, Upper Boundary is the higher boundary, Inferior Boundary is the unqualified boundary, and the boundary value exp(d i ) Please refer to IMO CII Guide G4.

[0072] S5. CII rating optimization model construction steps: Based on the calculated speed-optimized fuel consumption of the ship under empty and full-load states, the sailing distance of the ship under empty and full-load states, and the given annual ship CII rating, a Bayesian optimization algorithm is used to establish a CII rating optimization model with the ship CII rating as a constraint and maximizing the sailing distance of the ship under full load as the optimization goal.

[0073] Furthermore, this step adopts the Bayesian optimization algorithm, takes the average speeds of the three grades of ships, low, medium and high, and the respective sailing time proportions of the three grades of ships as decision variables, and the current annual ship CII value is less than or equal to the higher boundary value in the CII rating calculation rules, the average speed of the low-level ship is less than the 33% quantile of the historical daily speed, the average speed of the medium-level ship is between the 33% quantile and the 66% quantile of the historical daily speed, and the average speed of the high-level ship is between the 66% quantile of the historical daily speed and the maximum value of the historical daily speed, and establishes a CII rating optimization model with the ship CII rating as a constraint condition and maximizing the sailing distance of the ship under full load as the optimization goal.

[0074] Specifically, this step establishes an optimization model for ship navigation performance, economic benefit maximization, and CII rating constraints based on Bayesian optimization. The CII rating optimization model integrates the ship's CII compliance and economic maximization, and essentially includes establishing a constrained single-objective optimization model involving decision variables β i and V i .

[0075] (1) The optimization goal of the Bayesian optimization model is to maximize the fully loaded sailing distance of the ship.

[0076]

[0077] Among them, D T represents the total sailing distance for the whole year, p represents the ratio of full load to total sailing distance, T p It represents the full-load sailing time, i=1, 2, 3, β1, β2, β3 correspond to the low, medium and high speed sailing time of the ship respectively, and V1, V2, V3 correspond to the average speed of the ship at low, medium and high levels respectively.

[0078] (2) The constraints of the Bayesian optimization model are: CII rating constraints.

[0079]

[0080] Among them, Attained CII y ≤Required CII y *exp(d3)

[0081]

[0082] V1∈(0,V b1 ],V2∈(V b1 , V b2 ), V3∈(V b2 , V max ]

[0083]

[0084] Among them, V H represents the historical average speed (rounded), V b1 The 33% percentile of the historical daily speed, V b2 The 66% percentile of the historical daily speed, V max Indicates the maximum value of historical daily speed. Other parameters are the same as above.

[0085] S6, optimization result output step, after optimizing the CII rating of the ship according to the constructed CII rating optimization model, output the optimization results of the average speed of different speed levels and the optimization results of the sailing distance under empty and full load states.

[0086] Specifically, the ship's target CII rating is set based on the CII rating calculation rules, the ship's target CII rating is input into the established CII rating optimization model, and a set of feasible solutions is calculated and solved according to the preset initial value and number of iterations to complete the ship's CII rating optimization, and the average speed of the ship at different speed levels in the current year and the optimization results of the corresponding sailing time and the optimization results of the sailing distance under empty and full load conditions are output. In other words, after optimizing the ship's CII rating according to the Bayesian optimization algorithm, this step provides the optimization results of the average speed, sailing time, and full load sailing distance.

[0087] The specific method of each step is specifically introduced below in conjunction with the embodiment data.

[0088] S1: Multi-source data fusion step, selection of verification data.

[0089] Multi-source data such as AIS data, MRV data and meteorological data are selected for fusion and analysis to optimize the ship CII rating. In addition to basic ship information such as MMSI and date, the data set also mainly includes the daily sailing time, daily sailing distance and empty and full loading status of each ship. This implementation plan uses data from January 1, 2023 to December 31, 2023. A ship is selected for verification. To ensure data security and privacy, ship A is used here.

[0090] S2: Steps for constructing a ship fuel consumption calculation model to calculate the ship fuel consumption.

[0091] The total sailing fuel consumption of Ship A in 2023 is 28,405.38 tons. For heavy oil (heavy fuel oil), the carbon emission factor is 3.114; therefore, the carbon emissions of Ship A in 2023 are 88,454.35 tons.

[0092] S3: Speed ​​optimization step, determining the ship speed optimization model parameters.

[0093] The low speed, medium speed and high speed of ship A in 2023 are 2.64kn / h, 10.99kn / h and 18.96kn / h respectively, and the corresponding sailing time is 458.77h, 1242.30h and 4100.90h; the full-load sailing distance is 70763.28 nautical miles, and the annual average speed is 14.55kn / h. The ratio of the full-load sailing distance to the sailing distance is p=0.74.

[0094] After speed optimization, by calculating the empty and full-load fuel consumption of ship A, the total fuel consumption in 2023 is 21,075.76 tons.

[0095] S4: Ship CII rating calculation step, calculate the ship CII rating.

[0096] First, calculate the CII value of the ship in the calendar year 2023. The carbon emissions of the ship in 2023 calculated by S2 are 88454.35 tons. It can be seen that M = 88454.35 tons, the carrying capacity of ship A Capacity = 197021 tons, and the sailing distance of ship A in the calendar year 2023 Dt = 95619.13 nautical miles; therefore W = D t *Capacity,

[0097] Secondly, calculate the CII baseline. Ship A is a container ship, and the parameters a and c are 1984.00 and 0.489 respectively; the CII baseline is based on 2019, so the baseline reference value of Ship A in 2019 is CII ref =a*Capacity -c =1984*197021- 0.489 =5.11.

[0098] Next, calculate the Required CII of Ship A in the calendar year 2023 = CII ref (1-Z). The value of Z in 2023 (Z can be found in Table 1) is selected as 5%, so Required CII = 4.86.

[0099] Finally, the CII rating is calculated, According to the above CII rating calculation rules, Ship A's CII rating in 2023 is C.

[0100] S5: CII rating optimization model construction steps, and relevant parameter results through Bayesian optimization.

[0101] The Bayesian optimization algorithm is used to establish a CII rating optimization model with the ship's CII rating as a constraint and the maximization of the ship's sailing distance under full load as the optimization goal. According to IMO Guidelines G1 and G4, the DWT of each ship should be used as the capacity in the CII calculation formula, and the ratio of the ship's current CII to the required CII needs to be less than or equal to 0.83 to obtain a CII rating of A. This implementation plan sets the target CII rating of ship A in 2024 to be A, and the full-load sailing mileage in this implementation plan is not less than 95% of the full-load sailing mileage of ship A in 2023.

[0102] S6: Optimization result output step, outputting the optimization result.

[0103] As a set of feasible solutions obtained by the optimizer, the low speed, medium speed and high speed of ship A in 2024 are 14.15 kn / h, 18.92 kn / h and 22.18 kn / h respectively, and the corresponding sailing time is 4566.64h, 1800.57h and 0h; the full load sailing distance is 68799.73 nautical miles. According to this speed distribution, the annual CII can reach A rating. The optimization results are as follows Figure 2 As shown, it shows the optimization results such as low, medium and high speeds and sailing time, full-load voyage, total fuel consumption, total carbon emissions, CII value, CII grade and the ratio of full-load voyage in the previous year in each year from 2023 to 2026.

[0104] The present invention also relates to a ship CII rating optimization system based on Bayesian optimization, which corresponds to the above-mentioned ship CII rating optimization method based on Bayesian optimization, and can be understood as a system for implementing the above-mentioned ship CII rating optimization method based on Bayesian optimization, and its structure is as follows: Figure 3 As shown, it includes a multi-source data fusion module, a ship fuel consumption calculation model construction module, a speed optimization module, a ship CII rating calculation module, a CII rating optimization model construction module and an optimization result output module which are connected in sequence.

[0105] Among them, the multi-source data fusion module fuses and analyzes the ship AIS data and MRV message data from several data sources to form the basic data of the ship's historical operation; the ship fuel consumption calculation model construction module, based on the ship's historical operation basic data, combines the cubic relationship between the ship's fuel consumption and the ship's rotation speed from the MRV message data, establishes a ship fuel consumption calculation model, and calculates the ship's fuel consumption and ship carbon emissions according to the ship fuel consumption calculation model; the speed optimization module, based on the ship's historical operation basic data, divides the ship's historical navigation speed from the ship's AIS data into several speed levels, and calculates the average speed of different speed levels, and calculates the navigation time of the ship in the empty and full-load states by combining the proportion of the ship's sailing time at different speed levels, the proportion of the ship's empty and full-load state distance, and the average speed of different speed levels. distance, and convert the calculated ship fuel consumption into the fuel consumption of the ship in the speed-optimized empty and full-load states; the ship CII rating calculation module calculates the ship CII rating of a given year based on the historical operation basic data of the ship and the calculated ship carbon emissions; the CII rating optimization model construction module, according to the calculated speed-optimized ship fuel consumption in the empty and full-load states, the sailing distance of the ship in the empty and full-load states, and the given year ship CII rating, adopts the Bayesian optimization algorithm to establish a CII rating optimization model with the ship CII rating as a constraint and maximizing the sailing distance of the ship in the full-load state as the optimization goal; the optimization result output module, after optimizing the ship CII rating according to the constructed CII rating optimization model, outputs the optimization results of the average speed of different speed levels and the optimization results of the sailing distance in the empty and full-load states.

[0106] Preferably, in the ship fuel consumption calculation model construction module, based on the cubic relationship between the ship fuel consumption and the ship rotation speed derived from the MRV message data in the ship's historical operation basic data and the empty and full load states determined by the draft data derived from the ship's AIS data, a ship fuel consumption calculation model is established to calculate the ship's fuel consumption in the full load state and the ship's fuel consumption in the empty load state as well as the corresponding ship carbon emissions.

[0107] Preferably, in the speed optimization module, based on the historical operation basic data of the ship, the historical sailing speed of the ship derived from the ship AIS data is divided into three levels: low, medium and high, and the average speeds of the ships at the low, medium and high levels are calculated respectively; the sailing distance of the ship in the empty and full-load states is calculated by combining the respective sailing time proportions of the ships at the low, medium and high levels, the total sailing time throughout the year, the proportion of the distance of the ships in the empty and full-load states, and the average speeds of the ships at the low, medium and high levels; and according to the respective sailing time proportions of the ships at the low, medium and high levels, the total sailing time throughout the year, the proportion of the distance of the ships in the empty and full-load states, and the average speeds of the ships at the low, medium and high levels, the calculated fuel consumption of the ship in the full-load state and the fuel consumption of the ship in the empty-load state are respectively converted into the fuel consumption of the ship in the empty and full-load states with optimized speed of each level.

[0108] Preferably, in the ship CII rating calculation module, the product of the ship carrying capacity and the sailing distance in a given year is taken as the total transport work undertaken, and the ship CII value in the given year is calculated by the ratio of the calculated ship carbon emissions to the total transport work undertaken; and the curve of the median of the operational carbon intensity performance of the ship in a certain historical year is taken as the CII baseline reference value, and the required ship CII value in the given year is calculated according to the requirements of the International Maritime Organization on carbon emissions and in combination with the CII baseline reference value and the CII rating reduction factor; and then the ratio between the ship CII value in the given year and the required ship CII value is calculated and according to the CII rating calculation rules, the CII rating of the ship in the given year is obtained.

[0109] The invention discloses a ship CII rating optimization method and system based on Bayesian optimization, integrates several types of data such as ship AIS data, MRV message data, meteorological data, etc. to form ship historical operation data, provides a ship CII calculation and rating scheme based on historical operation data, redistributes the proportions of different speed ranges through speed optimization, adjusts the corresponding average values, establishes a speed optimization model for operating ships, and strikes a balance between ship operation economic maximization and environmental impact, adopts a scheme based on Bayesian optimization CII rating to maximize economic benefits, and meets the requirements of the International Maritime Organization on CII at the same time, and provides a navigation reference for the navigation speed of ships under empty and full-load states based on the ship CII rating based on Bayesian optimization.

[0110] It should be noted that the above-described specific implementations can enable those skilled in the art to more fully understand the invention, but do not limit the invention in any way. Therefore, although this specification has described the invention in detail with reference to the drawings and embodiments, those skilled in the art should understand that the invention can still be modified or replaced by equivalents. In short, all technical solutions and improvements that do not deviate from the spirit and scope of the invention should be included in the protection scope of the patent for the invention.

Claims

1. A ship CII rating optimization method based on Bayesian optimization, characterized in that: The following steps are involved: The multi-source data fusion step is to fuse and analyze the ship AIS data and MRV message data from several data sources to form the basic data of the ship's historical operation; A ship fuel consumption calculation model construction step, based on the ship's historical operation basic data, combined with the cubic relationship between the ship's fuel consumption and the ship's rotation speed from the MRV message data, to establish a ship fuel consumption calculation model, and calculate the ship's fuel consumption and the ship's carbon emissions according to the ship fuel consumption calculation model; a speed optimization step, based on the historical operation basic data of the ship, dividing the historical sailing speed of the ship from the ship AIS data into several speed levels, and calculating the average speed of different speed levels, combining the proportion of the sailing time of the ship at different speed levels, the proportion of the distance of the ship in the empty and full-load states, and the average speed of different speed levels to calculate the sailing distance of the ship in the empty and full-load states, and converting the calculated ship fuel consumption into the fuel consumption of the speed-optimized ship in the empty and full-load states; The ship CII rating calculation step is to calculate the ship CII rating for a given year based on the historical operation basic data of the ship and the calculated ship carbon emissions; The steps of constructing the CII rating optimization model are as follows: based on the calculated fuel consumption of the ship under empty and full load conditions with optimized speed, the sailing distance of the ship under empty and full load conditions, and the CII rating of the ship in a given year, a CII rating optimization model is established using the Bayesian optimization algorithm with the ship's CII rating as a constraint and the maximization of the sailing distance of the ship under full load conditions as the optimization goal; The optimization result output step optimizes the CII rating of the ship according to the constructed CII rating optimization model, and outputs the optimization results of the average speed of different speed levels and the optimization results of the sailing distance under empty and full load conditions.

2. The ship CII rating optimization method according to claim 1, characterized in that: In the ship fuel consumption calculation model construction step, based on the cubic relationship between the ship fuel consumption and the ship rotation speed derived from the MRV message data in the ship's historical operation basic data and the empty and full load states determined by the draft data, a ship fuel consumption calculation model is established and the ship fuel consumption in the full load state and the ship fuel consumption in the empty load state and the corresponding ship carbon emissions are calculated.

3. The ship CII rating optimization method according to claim 2, characterized in that: In the speed optimization step, based on the historical operation basic data of the ship, the historical navigation speed of the ship derived from the ship AIS data is divided into three levels: low, medium and high, and the average speeds of the ships at the low, medium and high levels are calculated respectively. The navigation distance of the ship in the empty and full-load states is calculated by combining the respective navigation time proportions of the ships at the low, medium and high levels, the total navigation time throughout the year, the proportion of the distance of the ships in the empty and full-load states, and the average speeds of the ships at the low, medium and high levels; and according to the respective navigation time proportions of the ships at the low, medium and high levels, the total navigation time throughout the year, the proportion of the distance of the ships in the empty and full-load states, and the average speeds of the ships at the low, medium and high levels, the calculated fuel consumption of the ship in the full-load state and the fuel consumption of the ship in the empty-load state are respectively converted into the fuel consumption of the ships in the empty and full-load states with optimized speeds of each level.

4. The ship CII rating optimization method according to any one of claims 1 to 3, characterized in that: In the ship CII rating calculation step, the product of the ship carrying capacity and the sailing distance in a given year is taken as the total transport work undertaken, and the ship CII value for the given year is calculated by the ratio of the calculated ship carbon emissions to the total transport work undertaken; and the curve of the median of the operational carbon intensity performance of the ship in a certain historical year is taken as the CII baseline reference value, and the ship CII value required for the given year is calculated according to the requirements of the International Maritime Organization on carbon emissions and in combination with the CII baseline reference value and the CII rating reduction factor; and then the ratio between the ship CII value for the given year and the required ship CII value is calculated and according to the CII rating calculation rules, the ship CII rating for the given year is obtained.

5. The ship CII rating optimization method according to claim 1, characterized in that: In the step of constructing the CII rating optimization model, a Bayesian optimization algorithm is used, and the average speeds of the three grades of ships, low, medium and high, and the respective sailing time proportions of the three grades of ships are taken as decision variables, and the current annual ship CII value is less than or equal to the higher boundary value in the CII rating calculation rule, the average speed of the low-level ship is less than the 33% quantile of the historical daily speed, the average speed of the medium-level ship is between the 33% quantile and the 66% quantile of the historical daily speed, and the average speed of the high-level ship is between the 66% quantile of the historical daily speed and the maximum value of the historical daily speed. A CII rating optimization model is established with the ship CII rating as a constraint condition and maximizing the sailing distance of the ship under full load as the optimization goal.

6. The ship CII rating optimization method according to claim 1, characterized in that: In the optimization result output step, the ship's target CII rating is set based on the CII rating calculation rules, the ship's target CII rating is input into the established CII rating optimization model, and a set of feasible solutions are calculated and solved according to the preset initial value and number of iterations to complete the ship's CII rating optimization, and the optimization results of the ship's average speed and corresponding sailing time at different speed levels in the current year and the optimization results of the sailing distance under empty and full load conditions are output.

7. A ship CII rating optimization system based on Bayesian optimization, characterized in that: It includes a multi-source data fusion module, a ship fuel consumption calculation model construction module, a speed optimization module, a ship CII rating calculation module, a CII rating optimization model construction module and an optimization result output module, which are connected in sequence. The multi-source data fusion module performs multi-source data fusion and analysis on the ship AIS data and MRV message data from several data sources to form basic data of historical ship operations; The ship fuel consumption calculation model building module establishes a ship fuel consumption calculation model based on the ship historical operation basic data and the cubic relationship between the ship fuel consumption and the ship rotation speed derived from the MRV message data, and calculates the ship fuel consumption and the ship carbon emissions according to the ship fuel consumption calculation model; The speed optimization module, based on the historical operation basic data of the ship, divides the historical sailing speed of the ship from the ship AIS data into several speed levels, calculates the average speed of different speed levels, calculates the sailing distance of the ship in the empty and full-load state by combining the proportion of the sailing time of the ship at different speed levels, the proportion of the distance of the ship in the empty and full-load state, and the average speed of different speed levels, and converts the calculated fuel consumption of the ship into the fuel consumption of the ship in the empty and full-load state with speed optimization; The ship CII rating calculation module calculates the ship CII rating for a given year based on the ship's historical operating basic data combined with the calculated ship carbon emissions; The CII rating optimization model building module uses a Bayesian optimization algorithm to establish a CII rating optimization model with the ship's CII rating as a constraint and maximizing the ship's sailing distance under a fully loaded state as an optimization goal, based on the calculated fuel consumption of the ship under the speed-optimized empty and full-load states, the sailing distance of the ship under the empty and full-load states, and the CII rating of the ship in a given year; The optimization result output module optimizes the CII rating of the ship according to the constructed CII rating optimization model, and outputs the optimization results of the average speed of different speed levels and the optimization results of the sailing distance under empty and full load states.

8. The ship CII rating optimization system according to claim 7, characterized in that: In the ship fuel consumption calculation model construction module, based on the cubic relationship between the ship fuel consumption and the ship rotation speed derived from the MRV message data in the ship's historical operation basic data and the empty and full load states determined by the draft data derived from the ship's AIS data, a ship fuel consumption calculation model is established to calculate the ship's fuel consumption in the full load state and the ship's fuel consumption in the empty load state as well as the corresponding ship carbon emissions.

9. The ship CII rating optimization system according to claim 8, characterized in that: In the speed optimization module, based on the historical operation basic data of the ship, the historical sailing speed of the ship from the ship AIS data is divided into three levels: low, medium and high, and the average speeds of the ships at the low, medium and high levels are calculated respectively. The sailing distance of the ship in the empty and full-load states is calculated by combining the respective sailing time proportions of the ships at the low, medium and high levels, the total sailing time throughout the year, the proportion of the distance of the ships in the empty and full-load states, and the average speeds of the ships at the low, medium and high levels; and according to the respective sailing time proportions of the ships at the low, medium and high levels, the total sailing time throughout the year, the proportion of the distance of the ships in the empty and full-load states, and the average speeds of the ships at the low, medium and high levels, the calculated fuel consumption of the ship in the full-load state and the fuel consumption of the ship in the empty-load state are respectively converted into the fuel consumption of the ships in the empty and full-load states with optimized speeds of each level.

10. The ship CII rating optimization system according to any one of claims 7 to 9, characterized in that: In the ship CII rating calculation module, the product of the ship's carrying capacity and the sailing distance in a given year is taken as the total transport work undertaken, and the ship's CII value for a given year is calculated by the ratio of the calculated ship's carbon emissions to the total transport work undertaken; and the curve of the ship reaching the median of the operational carbon intensity performance in a certain historical year is taken as the CII baseline reference value, and the ship's required CII value for the given year is calculated according to the requirements of the International Maritime Organization on carbon emissions and in combination with the CII baseline reference value and the CII rating reduction factor; and then the ratio between the ship's CII value for the given year and the required ship's CII value is calculated and according to the CII rating calculation rules, the ship's CII rating for the given year is obtained.

Images