Method and system for optimizing voyage number of new fuel dry bulk carrier route
The voyage planning of new fuel dry bulk carriers is optimized through the non-dominant sorting genetic algorithm, which solves the problems of traditional fuel ship pollution and inefficient route planning of new fuel ships, and improves operational efficiency and cost reduction, and reduces environmental pollution.
Patent Information
- Application Number
- CN202411037257.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-30
- Publication Date
- 2024-11-26
- Estimated Expiration
- 2044-07-30
AI Technical Summary
Traditional fuel ships generate a large amount of waste gas and pollutants during operation, which has a serious impact on the environment. The route planning of new fuel ships has problems such as inefficient operation and high cost.
Using non-dominant sorting genetic algorithm to optimize the voyage planning of new fuel dry bulk carriers, design customized voyage optimization by obtaining ship and navigation parameters, establishing comprehensive objective functions, maximizing the total transfer volume of dry bulk cargo and minimizing the total voyage cost, and designing customized voyage optimization Methods and systems.
It effectively improves operational efficiency, reduces operating costs, reduces no-load driving and waiting time, reduces energy consumption and environmental pollution, and maximizes the total transfer volume of dry bulk cargo and minimizes the total range cost.
Smart Images

Figure CN119026732A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of ship operation, and in particular to a voyage optimization method and system for a new fuel dry bulk carrier route. Background Art
[0002] Traditional fuel ships produce a large amount of exhaust gas and pollutants during operation, which has a serious impact on the environment. The fuel used by traditional fuel ships, such as heavy oil or diesel, releases a large amount of carbon dioxide (CO2), nitrogen oxides (NOx) and sulfur oxides (SOx) when burned. These exhaust gases are major greenhouse gases and air pollutants, which have a direct impact on climate change and air quality. In addition, the combustion of traditional fuel ships also produces other chemical pollutants, such as carbon monoxide (CO), volatile organic compounds (VOCs) and particulate matter (PM). These pollutants are harmful to air quality and human health, especially fine particulate matter (PM2.5), which can cause respiratory diseases and cardiovascular problems. In addition, traditional fuel ships produce a large amount of wastewater during navigation, including ship drainage, washing water and toilet sewage. These wastewaters contain petroleum residues, heavy metals and other harmful substances. If they are not properly treated and discharged directly into the ocean, they will have a serious impact on aquatic life and marine ecosystems. The engines and propellers of traditional fuel ships generate noise pollution when they are running, which disturbs and damages marine life, especially aquatic animals. Noise pollution can affect the migration, communication and reproduction of aquatic organisms, and have an adverse effect on the ecological balance. Therefore, in order to reduce the impact of traditional fuel ships on the environment, it is very important to promote and apply new fuel technologies, adopt advanced exhaust gas treatment equipment, implement effective oil-water separation and wastewater treatment, and control noise. The introduction of new fuel ships, such as liquefied natural gas (LNG) ships and electric ships, can significantly reduce pollutant emissions and noise pollution, which is of positive significance to protecting the environment.
[0003] However, there are obvious differences between the operation of new fuel ships and traditional ships. For example, changes in fuel filling points will lead to adjustments in routes, and changes in emission restriction areas also need to be considered. Therefore, in order to improve operational efficiency and reduce costs, how to reasonably formulate routes has become an important research topic. First, changes in fuel filling points will directly affect route planning. New fuel ships may need to refuel more frequently at specific ports, so route planning needs to consider the location and time of fuel filling points. This requires route planners to fully understand the energy needs of new fuel ships and comprehensively consider the feasibility of port facilities and supply chains. Secondly, with the improvement of environmental protection requirements, the scope and restrictions of emission restriction areas are also changing. Route planning must comply with relevant emission restriction regulations and arrange routes reasonably to avoid or minimize entry into emission restriction areas. This requires timely acquisition of the latest emission restriction area information and incorporating it into the considerations of route planning. In order to improve operational efficiency and reduce costs, it is crucial to reasonably formulate routes. Route planning needs to take into account multiple factors, including voyage distance, sea conditions, traffic flow, port facilities and navigation safety. Reasonable route planning can reduce ship sailing distance and time, reduce fuel consumption and operating costs. At the same time, taking into account the requirements of environmental protection, route planning should also try to avoid entering sensitive ecological areas and reduce the impact on marine life.
[0004] After the route planning is determined, formulating a reasonable annual sailing plan for new fuel ships is a key step to optimize the operation of the fleet, reduce costs and increase cargo transshipment. In order to achieve this goal, the method of operations optimization can be used to reasonably allocate the fleet through network optimization scheduling. Network optimization scheduling is a complex problem that requires multiple factors to make the best decision. First of all, the ship's loading capacity, voyage time and port docking time are all factors that need to be considered. The ship's loading capacity directly affects the amount of cargo that can be carried on each voyage, while the voyage time and port docking time affect the ship's transportation efficiency and time cost. In addition, fluctuating loading information is also a very important consideration. By timely understanding the supply situation at the terminal, the fleet's operation strategy can be better adjusted. For example, if the port is in the rainy season and the loading time is long, the ship's sailing time can be increased to meet the port's loading and unloading capacity, and fuel consumption can also be reduced, reducing costs; and if the loading speed in a certain season is faster, the voyage can be appropriately accelerated to increase the terminal's output. In order to achieve efficient fleet operations, intelligent scheduling algorithms are crucial tools. Through network optimization scheduling and intelligent scheduling algorithms, the fleet can achieve efficient operation and improve overall transportation efficiency. This will bring many benefits, including reducing operating costs, increasing cargo transfer volume, reducing empty driving and waiting time, and reducing energy consumption and environmental pollution. The use of intelligent scheduling algorithms can help achieve efficient operation of the fleet, maximize benefits and achieve improved transportation efficiency. In short, from the perspective of operational optimization, after the route planning is determined, it is very important for the fleet to formulate a reasonable annual sailing plan. Summary of the invention
[0005] In order to solve the problems of low operational efficiency and high cost in the current voyage optimization process, the present invention provides a voyage optimization method for a new fuel dry bulk carrier route, with the goal of maximizing the total dry bulk cargo transshipment volume and minimizing the total voyage cost, and with the value range of the decision variable, the limited range of the fleet size, the minimum number of voyages, the limited range of the cumulative loading volume of the ship at the loading port, etc. as constraints, establish the first and second objective functions, and then construct a comprehensive objective function, and design a customized non-dominated sorting genetic algorithm to obtain the optimal solution of the comprehensive objective function, so as to maximize the total dry bulk cargo transshipment volume and minimize the total voyage cost, which can effectively improve operational efficiency and reduce costs. The present invention also relates to a voyage optimization system for a new fuel dry bulk carrier route.
[0006] The technical solution of the present invention is as follows:
[0007] A method for optimizing the voyage of a new fuel dry bulk carrier route, characterized in that it comprises the following steps:
[0008] Parameter acquisition step: acquiring ship parameters and navigation parameters of the new fuel dry bulk carrier, wherein the ship parameters include the maximum loading capacity of the ship, the fleet size, the minimum number of voyages, and the cumulative loading capacity of the ship at the loading port; the navigation parameters include the starting time of the voyage, the duration of the voyage, and the cost of the voyage;
[0009] Objective function establishment step: constructing decision variables based on the judgment result of whether the ship starts sailing at the starting sailing time of the voyage and the sailing duration, and establishing a first objective function according to the maximum loading capacity of the ship and the decision variables under multiple constraints established by the value range of the decision variables, the limit range of the fleet size, the minimum number of voyages, and the limit range of the cumulative loading volume of the ship at the loading port as constraints, and establishing a second objective function according to the voyage cost determined by the starting sailing time and the sailing duration of the voyage and the decision variables, and then merging the first objective function and the second objective function based on the preset weight coefficient and the weighted sum method to obtain a comprehensive objective function;
[0010] Initial population generation step: the total sailing time of the ship is divided into multiple time periods in chronological order to generate the initial population, and each time period in the initial population is regarded as an individual;
[0011] Fitness value calculation and offspring individual generation step: the fitness value of each individual in the initial population is calculated according to the first objective function value and its weight coefficient, and the second objective function value and its weight coefficient, and the genetic operator in the non-dominated sorting genetic algorithm is used to perform neighborhood search evolution operation on multiple individuals in the initial population to generate multiple new offspring individuals, and the fitness value of each new offspring individual is calculated;
[0012] New population generation step: merge the new offspring individuals with the parent individuals to generate a temporary population, and compare the number of individuals in the temporary population with the preset number threshold. If the number of individuals in the temporary population is greater than the preset number threshold, sort the fitness value of each individual in the temporary population by size, and delete the individuals that exceed the number threshold in the sorted temporary population from small to large to obtain a new population;
[0013] Optimal solution calculation steps: Based on the new population, the fitness value calculation and offspring individual generation steps and the new population generation steps are repeated until the preset number of iterations is reached. The individual with the smallest fitness value in the new population when the preset number of iterations is reached is taken as the optimal solution of the comprehensive objective function, thereby maximizing the total dry bulk cargo transshipment volume and minimizing the total voyage cost.
[0014] Preferably, in the fitness value calculation and offspring individual generation steps, the genetic operators include a selection operator, a new blood operator, a crossover operator, an insertion operator and a mutation operator; using the genetic operators in the non-dominated sorting genetic algorithm to perform neighborhood search evolution operations on multiple individuals in the initial population to generate multiple new offspring individuals specifically includes:
[0015] S1: The selection operator selects multiple individuals as parent individuals from the initial population according to the fitness value and adopts the roulette wheel selection method;
[0016] S2: The new blood operator randomly generates multiple new individuals and adds them to the parent individuals;
[0017] S3: Determine whether to perform a crossover operation based on the randomly generated probability value and the preset crossover rate. If yes, the crossover operator randomly selects two individuals from the parent individuals, and then randomly selects two time units with relative positions from the two individuals, and exchanges the codes of the two time units to generate two offspring individuals; if no, the number of parent individuals remains unchanged, and all individuals in the parent individuals are used as offspring individuals;
[0018] S4: Repeat step S3 until the preset number of iterations is reached to obtain multiple offspring individuals;
[0019] S5: The insertion operator performs an insertion operation on each offspring individual to obtain multiple inserted offspring individuals;
[0020] S6: The mutation operator performs mutation operations on each offspring individual after insertion to obtain multiple new offspring individuals.
[0021] Preferably, in the parameter acquisition step, port parameters are also acquired, and the port parameters include the initial inventory of the loading port and the loading capacity of the loading port at a certain moment.
[0022] Preferably, in the objective function establishment step, the multiple constraints include a first constraint, a second constraint, a third constraint and a fourth constraint, the first constraint is constructed based on the decision variable and a preset numerical threshold, the second constraint is constructed based on the decision variable and the fleet size, the third constraint is constructed based on the decision variable and the minimum number of voyages, and the fourth constraint is constructed based on the cumulative loading volume of the ship at the loading port and the maximum loading volume of the ship.
[0023] Preferably, in the objective function establishment step, the multiple constraints also include a fifth constraint, a sixth constraint, a seventh constraint and an eighth constraint, the fifth constraint is constructed according to the cumulative loading volume of the ship at the loading port, the initial inventory of the loading port, the loading capacity of the loading port at a certain moment, the maximum loading volume of the ship and the decision variables, the sixth constraint is constructed according to the cumulative loading volume of the ship at the loading port, the loading capacity of the loading port at a certain moment, the maximum loading volume of the ship and the decision variables, the seventh constraint is constructed according to the initial inventory of the loading port, the maximum loading volume of the ship and the decision variables, and the eighth constraint is constructed according to the cumulative loading volume of the ship at the loading port, the maximum loading volume of the ship and the decision variables.
[0024] Preferably, in the initial population generation step, each time period is divided into a plurality of time units in chronological order, and each time unit is encoded to obtain the encoded time units.
[0025] A voyage optimization system for a new fuel dry bulk carrier route, characterized by comprising a parameter acquisition module, an objective function establishment module, an initial population generation module, a fitness value calculation and offspring individual generation module, a new population generation module and an optimal solution calculation module connected in sequence,
[0026] The parameter acquisition module acquires the ship parameters and navigation parameters of the new fuel dry bulk carrier, wherein the ship parameters include the maximum loading capacity of the ship, the fleet size, the minimum number of voyages, and the cumulative loading capacity of the ship at the loading port; the navigation parameters include the starting time of the voyage, the duration of the voyage, and the cost of the voyage;
[0027] The objective function establishment module constructs decision variables based on the judgment result of whether the ship starts sailing at the starting sailing time of the voyage and the sailing duration, and establishes a first objective function according to the maximum loading capacity of the ship and the decision variables under multiple constraint conditions established by the value range of the decision variables, the limit range of the fleet size, the minimum number of voyages, and the limit range of the cumulative loading volume of the ship at the loading port as constraints, and establishes a second objective function according to the voyage cost determined by the starting sailing time and the sailing duration of the voyage and the decision variables, and then merges the first objective function and the second objective function based on a preset weight coefficient and a weighted sum method to obtain a comprehensive objective function;
[0028] The initial population generation module divides the total sailing time of the ship into multiple time periods in chronological order to generate an initial population, and regards each time period in the initial population as an individual;
[0029] The fitness value calculation and offspring individual generation module calculates the fitness value of each individual in the initial population according to the first objective function value and its weight coefficient, and the second objective function value and its weight coefficient, and uses the genetic operator in the non-dominated sorting genetic algorithm to perform a neighborhood search evolution operation on multiple individuals in the initial population to generate multiple new offspring individuals, and calculates the fitness value of each new offspring individual;
[0030] The new population generation module merges the new offspring individuals with the parent individuals to generate a temporary population, and compares the number of individuals in the temporary population with a preset number threshold. If the number of individuals in the temporary population is greater than the preset number threshold, the fitness value of each individual in the temporary population is sorted according to size, and the individuals exceeding the number threshold in the sorted temporary population are deleted in order from small to large to obtain a new population;
[0031] The optimal solution calculation module is based on the new population and repeats the fitness value calculation and the work of the offspring individual generation module and the new population generation module until a preset number of iterations is reached. The individual with the smallest fitness value in the new population when the preset number of iterations is reached is taken as the optimal solution of the comprehensive objective function, thereby maximizing the total dry bulk cargo transshipment volume and minimizing the total voyage cost.
[0032] Preferably, in the fitness value calculation and offspring individual generation module, the genetic operators include a selection operator, a new blood operator, a crossover operator, an insertion operator and a mutation operator; using the genetic operators in the non-dominated sorting genetic algorithm to perform neighborhood search evolution operations on multiple individuals in the initial population to generate multiple new offspring individuals specifically includes:
[0033] S1: The selection operator selects multiple individuals as parent individuals from the initial population according to the fitness value and adopts the roulette wheel selection method;
[0034] S2: The new blood operator randomly generates multiple new individuals and adds them to the parent individuals;
[0035] S3: Determine whether to perform a crossover operation based on the randomly generated probability value and the preset crossover rate. If yes, the crossover operator randomly selects two individuals from the parent individuals, and then randomly selects two time units with relative positions from the two individuals, and exchanges the codes of the two time units to generate two offspring individuals; if no, the number of parent individuals remains unchanged, and all individuals in the parent individuals are used as offspring individuals;
[0036] S4: Repeat step S3 until the preset number of iterations is reached to obtain multiple offspring individuals;
[0037] S5: The insertion operator performs an insertion operation on each offspring individual to obtain multiple inserted offspring individuals;
[0038] S6: The mutation operator performs mutation operations on each offspring individual after insertion to obtain multiple new offspring individuals.
[0039] Preferably, the parameter acquisition module also acquires port parameters, which include the initial inventory of the loading port and the loading capacity of the loading port at a certain moment.
[0040] Preferably, in the objective function establishment module, the multiple constraints include a first constraint, a second constraint, a third constraint, a fourth constraint, a fifth constraint, a sixth constraint, a seventh constraint and an eighth constraint; the first constraint is constructed according to a decision variable and a preset numerical threshold, the second constraint is constructed according to a decision variable and a fleet size, the third constraint is constructed according to a decision variable and a minimum number of voyages, the fourth constraint is constructed according to the cumulative loading of the ship at the loading port and the maximum loading of the ship, the fifth constraint is constructed according to the cumulative loading of the ship at the loading port, the initial inventory of the loading port, the loading capacity of the loading port at a certain time, the maximum loading of the ship and the decision variables, the sixth constraint is constructed according to the cumulative loading of the ship at the loading port, the loading capacity of the loading port at a certain time, the maximum loading of the ship and the decision variables, the seventh constraint is constructed according to the initial inventory of the loading port, the maximum loading of the ship and the decision variables, and the eighth constraint is constructed according to the cumulative loading of the ship at the loading port, the maximum loading of the ship and the decision variables.
[0041] The beneficial effects of the present invention are:
[0042] The present invention provides a voyage optimization method for a new fuel dry bulk carrier route. The method first obtains the ship parameters and navigation parameters of the new fuel dry bulk carrier, constructs decision variables based on the judgment result of whether the ship starts sailing at the starting sailing time of the voyage and the sailing duration, and establishes a first objective function according to the maximum loading capacity of the ship and the decision variables under multiple constraint conditions established by using the value range of the decision variables, the limit range of the fleet size, the minimum number of voyages, and the limit range of the cumulative loading capacity of the ship at the loading port as constraints, and establishes a second objective function according to the voyage cost determined by the starting sailing time and the sailing duration of the voyage and the decision variables, and then merges the first objective function and the second objective function based on a preset weight coefficient and a weighted sum method to obtain a comprehensive objective function; then divides the total sailing time of the ship into multiple time periods in chronological order to generate an initial population, and each time period in the initial population is divided into a plurality of time periods. Each time period is regarded as an individual, and then the fitness value of each individual in the initial population is calculated, and the genetic operator in the non-dominated sorting genetic algorithm is used to perform neighborhood search evolution operations on multiple individuals in the initial population to generate multiple new offspring individuals, and the new offspring individuals are merged with the parent individuals to generate a temporary population, and the number of individuals in the temporary population is compared with a preset number threshold, and the individuals exceeding the number threshold in the sorted temporary population are deleted to obtain a new population, and finally based on the new population, the fitness value calculation and the offspring individual generation steps and the new population generation steps are repeated until the preset number of iterations is reached, and the individual with the largest fitness value in the new population when the preset number of iterations is reached is taken as the optimal solution of the comprehensive objective function, thereby maximizing the total transshipment volume of dry bulk cargo (such as coal, ore, wood and other bulk dry bulk cargoes) and minimizing the total voyage cost, effectively improving operational efficiency and reducing costs.
[0043] The present invention optimizes the scheduling of the new fuel dry bulk carrier network under different speeds and different fuel filling strategies by considering factors such as multiple ship parameters and navigation parameters, so as to better adjust the fleet's operating strategy, thereby reducing fuel consumption and reducing costs. It also establishes a comprehensive objective function, which can also be called a dual-objective mixed integer programming function, by adopting a specific method, and designs a customized non-dominated sorting genetic algorithm to obtain the optimal solution of the comprehensive objective function, thereby realizing intelligent scheduling of new fuel dry bulk carriers, so that the fleet can achieve efficient operation and improve overall transportation efficiency, which will bring many benefits, including reducing operating costs, increasing dry bulk cargo transshipment volume, reducing empty driving and waiting time, and reducing energy consumption and environmental pollution. At the same time, it can also help the fleet achieve efficient operation, maximize benefits and improve transportation efficiency.
[0044] The present invention also relates to a voyage optimization system for a new fuel dry bulk carrier route, which corresponds to the above-mentioned voyage optimization method for the new fuel dry bulk carrier route, and can be understood as a system for realizing the above-mentioned voyage optimization method for the new fuel dry bulk carrier route, including a parameter acquisition module, an objective function establishment module, an initial population generation module, a fitness value calculation and offspring individual generation module, a new population generation module and an optimal solution calculation module connected in sequence, and each module works in coordination with each other, which can be understood as studying the new fuel ship operation deduction intelligent scheduling algorithm, developing an intelligent scheduling decision module, and performing optimal solution calculations on different speeds and different fuel filling strategies. The new fuel dry bulk carrier network under the strategy is optimized and scheduled, with the goal of maximizing the total dry bulk cargo transshipment volume and minimizing the total voyage cost. The first and second objective functions are established with the value range of decision variables, the limit range of fleet size, the minimum number of voyages, and the limit range of the cumulative loading volume of ships at the loading port as constraints. Then, a comprehensive objective function is constructed, and a customized non-dominated sorting genetic algorithm is designed to obtain the optimal solution of the comprehensive objective function, so as to maximize the total dry bulk cargo transshipment volume while minimizing the total voyage cost, and realize intelligent scheduling of new fuel dry bulk carriers, which can effectively improve operational efficiency and reduce costs. BRIEF DESCRIPTION OF THE DRAWINGS
[0045] Figure 1 It is a flow chart of the voyage optimization method of the new fuel dry bulk carrier route of the present invention.
[0046] Figure 2 It is a flow chart of the non-dominated sorting genetic algorithm of the present invention. DETAILED DESCRIPTION
[0047] The present invention will be described below in conjunction with the accompanying drawings.
[0048] Fleet operators need to arrange a sailing plan for the entire fleet every year. Assuming that the optimal refueling strategy for the new fuel ship on each transport voyage has been obtained, the fleet's sailing plan for the whole year needs to be scheduled and arranged due to limitations such as fleet size, seasonal fluctuations in loading rates at loading ports, and the capacity of loading platforms at loading ports. From the perspective of fleet operators, more dry bulk cargo transshipment needs to be transported at the lowest possible cost, so minimizing fleet operating costs and maximizing dry bulk cargo transshipment are dual decision-making goals. New fuel dry bulk carriers transport bulk dry bulk cargoes such as coal, ore, and timber. The following takes ore as an example, that is, minimizing fleet operating costs and maximizing ore transshipment are dual decision-making goals.
[0049] The present invention relates to a voyage optimization method for a new fuel dry bulk carrier route. The flow chart of the method is as follows: Figure 1 As shown, the following steps are included in sequence:
[0050] 1. Parameter acquisition step: Obtain the ship parameters and navigation parameters of the new fuel dry bulk carrier. The ship parameters include the maximum loading capacity of the ship, fleet size, minimum number of voyages, and the cumulative loading capacity of the ship at the loading port; the navigation parameters include the starting time of the voyage, the duration of the voyage, and the cost of the voyage. Furthermore, the port parameters are also obtained, including the initial inventory of the loading port and the loading capacity of the loading port at a certain time.
[0051] 2. Objective function establishment steps: Based on the judgment result of whether the ship starts sailing at the starting sailing time of the voyage and the duration of the voyage, a decision variable is constructed, and under multiple constraints established by the value range of the decision variable, the limit range of the fleet size, the minimum number of voyages, and the limit range of the cumulative loading volume of the ship at the loading port, a first objective function is established according to the maximum loading volume of the ship and the decision variable, and a second objective function is established according to the voyage cost determined by the starting sailing time and the duration of the voyage and the decision variable, and then based on the preset weight coefficient and using the weighted sum method, the first objective function and the second objective function are combined to obtain a comprehensive objective function. Further, the value range of the decision variable includes the decision variable and the preset numerical threshold, the limit range of the fleet size includes the decision variable and the fleet size, and the limit range of the cumulative loading volume of the ship at the loading port includes the cumulative loading volume of the ship at the loading port and the maximum loading volume of the ship.
[0052] Specifically, the goal is to maximize the total ore transfer volume and minimize the total voyage cost. First, the decision variables are constructed based on the judgment result of whether the ship starts sailing at the starting time of the voyage and the duration of the voyage. Under the constraints of the decision variable value range, fleet size limit, minimum number of voyages, and cumulative cargo volume limit of the ship at the loading port, the maximum loading volume Q of the ship and the decision variable Q are used as the constraints. Establish the first objective function f1 and express it as follows:
[0053]
[0054] In the above formula, Whether the ship starts sailing at the starting sailing time t of the voyage and the sailing duration is p, if so, then is 1, if not, then is 0, Q is the maximum load of the ship (also known as the maximum loading capacity of the ship), T is the time range of the voyage, and p is the duration of the voyage, which is determined according to the longest and shortest voyage time.
[0055] Then the voyage cost is determined based on the voyage start time and voyage duration. Joint decision variables The second objective function f2 is established and expressed as follows:
[0056]
[0057] In the above formula, It represents the cost of a voyage with a starting time of t and a voyage duration of p.
[0058] Among them, the multiple constraints include the first constraint, according to the decision variable The first constraint condition ensures that at most one voyage departs in each time period, which is expressed as follows:
[0059]
[0060] The second constraint is based on the decision variables The second constraint condition is to limit the number of ships in the same time period to the range of fleet size, which is expressed as follows:
[0061]
[0062] In the above formula, Indicates whether there is a The number of departures and voyage duration is p, N is the fleet size, d p b) the possible duration of the voyage; ti It represents the processing time before the ship starts sailing in time period ti.
[0063] The third constraint is based on the decision variables The third constraint is the minimum number of voyages required, which must meet the minimum profit value of each voyage and is expressed as follows:
[0064]
[0065] In the above formula, K is the minimum number of flights.
[0066] The fourth constraint is constructed based on the cumulative loading volume of the ship at the loading port and the maximum loading capacity of the ship. The fourth constraint represents the capacity limit of the loading port, that is, the loading port can only serve one ship at a time, which is expressed as follows:
[0067]
[0068] In the above formula, y t represents the cumulative loading volume of the ship at the loading port, and y t ≥0, becomes 0 when the ship leaves the port, Q represents the maximum loading capacity (loading capacity) of the ship.
[0069] Furthermore, the fifth constraint condition is based on the cumulative loading volume y of the ship at the loading port t , the initial inventory h0 of the loading port, the loading capacity l of the loading port at a certain time tn tn , the maximum loading capacity Q of the ship and the decision variables Construction; The sixth constraint condition is based on the cumulative loading volume y of the ship at the loading port t , the loading capacity of the loading port at a certain time l tn , the maximum loading capacity Q of the ship and the decision variables The fifth and sixth constraints follow the real loading logic that the cumulative loading volume gradually increases with the real-time loading capacity. The fifth constraint is as shown in the following formula (7), and the sixth constraint is as shown in the following formula (8):
[0070]
[0071] Furthermore, the seventh constraint condition is based on the initial inventory h0 of the loading port, the maximum loading capacity Q of the ship, and the decision variable Construction; The eighth constraint condition is based on the cumulative loading volume y of the ship at the loading port t , the maximum loading capacity Q of the ship and the decision variables The seventh and eighth constraints ensure that the voyage can only start when the loading is completed. The seventh constraint is as shown in the following formula (9), and the eighth constraint is as shown in the following formula (10):
[0072]
[0073] Furthermore, the domain of the decision variable can be shown as (11) and (12):
[0074]
[0075] Then, since the problem is multi-objective, the general solution method cannot be directly solved. Therefore, based on the preset weight coefficient, the first objective function f1 and the second objective function f2 are combined by the weighted sum method that combines the two objective functions together to obtain a comprehensive objective function, which is expressed as follows:
[0076] minλ·f2-(1-λ)·f1 (13)
[0077] In the above formula, λ is a preset weight coefficient.
[0078] Considering that a ship can only start the next voyage after the loading work of the previous voyage is completed, two constraints are set for the comprehensive objective function to ensure that the loading and unloading processes between consecutive voyages will not overlap or conflict, thereby ensuring the smooth progress of the transportation process. As shown in the following effective inequalities (14) and (15), they are respectively used as two constraints. This is achieved by ensuring the minimum loading and unloading time between two consecutive voyages. For example, constraint (14) stipulates that the start time of the new voyage should be greater than or equal to the start time of the previous voyage plus the time required for loading, and the definition of constraint (15) is similar.
[0079]
[0080] In the above formula, f t0 represents the loading time required to start loading from time t0, b ti It represents the processing time before the ship starts sailing in time period ti.
[0081] 3. Initial population generation step: The total sailing time of the ship is divided into multiple time periods in chronological order to generate the initial population, and each time period in the initial population is regarded as an individual. Furthermore, each time period is divided into multiple time units in chronological order, and each time unit is encoded to obtain the encoded time units.
[0082] Compared with single-objective programming problems, which usually have only one exact optimal solution, multi-objective programming problems have a set of Pareto optimal solutions. In order to obtain as many Pareto optimal solutions as possible within a reasonable computing time, the comprehensive objective function is solved based on the non-dominated sorting genetic algorithm (NSGA-II) to obtain the optimal solution when the preset number of iterations is reached. The overall flow chart of the NSGA-II algorithm is shown in the figure. Figure 2 shown.
[0083] Specifically, first, the total sailing time of the ship is divided into multiple time periods in chronological order to form an initial population, and each time period in the initial population is regarded as an individual. Then, each individual in the initial population is divided into multiple time units in chronological order, and each time unit is encoded, and the number of the initial population is set. Among them, the encoding includes a description of the start time and sailing duration of the voyage in each time unit. By encoding each time unit, a time series can be formed to clearly represent the voyage arrangement. The initial population can be understood as the beginning of the NSGA-II algorithm, which affects the iterative evolution and convergence effect of the NSGA-II algorithm. The random sequence can be generated and the number of the initial population can be set. Set each time period in the initial population as S=(s1,s2,…,s i ,…,s T), S represents a time series with T elements (i.e., time units), where s i ≠-1 indicates the duration of the voyage starting at the i-th time unit, s i =-1 means that no voyage starts in this time unit. As shown in Table 1, assuming that an individual has 10 time units (the length of the time unit is set to one day in this study, and can also be reasonably set according to the actual needs for the time precision of the voyage plan, such as 2 days or 0.5 days, etc., which is not uniquely limited here), then the initial population plans two voyages. The first voyage is a non-negative unit in s3 (i.e., the third time unit), indicating that if the voyage duration set is [15,30,45] (indicating that the allowed voyage duration can be 15 days, 30 days, and 45 days), there is a voyage starting from time unit 3 and returning after 30 time units (i.e., 30 days, here the voyage duration is selected as 30 days). The next voyage is in s8 (i.e., the eighth time unit), indicating that the voyage starts from the eighth time unit and returns after 15 time units (i.e., 15 days, here the voyage duration is selected as 15 days).
[0084] Table 1
[0085] -1 -1 2 -1 -1 -1 -1 1 -1 -1
[0086] It should be noted that each individual (also called region) is improved independently, and each region is regarded as a self-individual, and the self-individual is continuously updated in the subsequent improvement search process. In addition, since the NSGA-II algorithm introduces random factors to achieve the purpose of continuous iterative updates, random factors can lead to local optimality. Therefore, the setting of this isolation area can also avoid the occurrence of this problem and try to ensure stable search in each area. In addition, after encoding each time unit, the encoding needs to be evaluated. The fitness value can be used to evaluate the quality of the individual, and the fitness value plays a decisive role in the algorithm.
[0087] 4. Fitness value calculation and offspring individual generation step: The fitness value of each individual in the initial population is calculated according to the first objective function value and its weight coefficient, and the second objective function value and its weight coefficient, and the genetic operator in the non-dominated sorting genetic algorithm is used to perform neighborhood search evolution operation on multiple individuals in the initial population to generate multiple new offspring individuals, and the fitness value of each new offspring individual is calculated. Preferably, the genetic operator includes a selection operator, a new blood operator, a crossover operator, an insertion operator and a mutation operator.
[0088] Specifically, the fitness value of each individual in the initial population is first calculated based on the first objective function value and its weight coefficient, and the second objective function value and its weight coefficient, and then a variety of genetic operators are designed to improve and upgrade the individuals in the initial population. The genetic operators may include selection operators, new blood operators, crossover operators, insertion operators and mutation operators, among which,
[0089] 1. Selection operator: It is the beginning of the evolution operator. It plays a key role in maintaining the continuous inheritance of good individuals and temporarily delaying the spread of poor individuals during the evolution process. The selection operator selects multiple individuals from the initial population as parent individuals based on the fitness value and the roulette selection method. The specific steps are as follows:
[0090] 1) Calculate the cumulative probability value of each individual according to the proportion of the individual fitness value in the overall total fitness value (the total fitness value is the sum of the fitness values of all individuals);
[0091] 2) Generate a random number in the interval [0,1] and use it as the discrimination value to determine the selected individual. If the random number falls within the cumulative probability value interval of an individual, the individual is selected;
[0092] 3) Add the selected individuals to the candidate improvement set, and repeat the above steps continuously until the number of candidate improvement sets reaches a certain number, and use the candidate improvement set as the parent individual.
[0093] 2. Crossover operator: This is to randomly cross the two annual sailing plans of the fleet. The specific steps are as follows:
[0094] 1) First, a probability value is randomly generated (usually a random floating point number between 0 and 1), and then the probability value is compared with the preset crossover rate; when the generated probability value is less than the crossover rate, the crossover operation is selected; when the generated probability value is greater than the crossover rate, the crossover step is directly skipped, the number of parent individuals is kept unchanged, and all individuals in the parent individuals are used as offspring individuals;
[0095] 2) After determining to perform the crossover operation, randomly select two individuals from the parent individuals;
[0096] 3) Randomly select two time units with relative positions from two individuals, and exchange the codes of these two time units to generate two offspring individuals; for example, as shown in Table 2, the code of the eighth time unit is selected for exchange, the eighth time unit of code 1 changes from "1" to "-1", and the eighth time unit of code 2 changes from "-1" to "1";
[0097] 4) Repeat steps 2 and 3 until the preset number of iterations is reached to obtain multiple offspring individuals.
[0098] Table 2
[0099]
[0100] It should be noted that the above crossover rate is a set value between 0 and 1, which indicates the probability of performing a crossover operation. For example, if the set crossover rate is 0.8, then in each crossover operation, there is an 80% probability of performing a crossover, and a 20% probability of not performing a crossover, and if a crossover is not performed, the parent individual is directly copied to the child individual.
[0101] 3. Insertion operator: Perform an insertion operation on each offspring individual to obtain multiple offspring individuals after insertion. That is, randomly select a time unit at a certain position from an offspring individual, and randomly insert the code of the time unit before the time unit at another position. As shown in Table 3, the code -1 of the fifth time unit is inserted before the code 2 of the third time unit to form an offspring individual after insertion.
[0102] Table 3
[0103]
[0104] 4. Mutation operator: Perform mutation operation on each inserted offspring individual to obtain multiple mutated new offspring individuals, which can prevent the algorithm from falling into local optimality and increase regional diversity, which is a key factor for the algorithm to conduct global exploration. That is, randomly select two time units from a new offspring individual and exchange the codes of the time units at these two positions. As shown in Table 4, the code -1 of the fifth time unit is exchanged with the code 2 of the third time unit to obtain a mutated new offspring individual. Then calculate the fitness value of each new offspring individual.
[0105] Table 4
[0106]
[0107] 5. New blood operator: In order to ensure the diversity of regional individual evolution, the new blood operator is selected to randomly generate multiple new individuals and add them to the parent individuals.
[0108] 5. New population generation step: merge the new offspring individuals with the parent individuals to generate a temporary population. In order to avoid the slow calculation caused by the increasing number of individuals in the temporary population, the number of individuals in the temporary population is compared with the preset number threshold. If the number of individuals in the temporary population is greater than the preset number threshold, the fitness value of each individual in the temporary population is sorted by size, and the individuals exceeding the number threshold in the sorted temporary population are deleted in order from small to large to obtain a new population. This step can be understood as setting a size-maintaining operator for the region. When the number of regions reaches a certain number, they are arranged from small to large according to the fitness value, and the redundant individuals at the end are deleted.
[0109] 6. Optimal solution generation step: Based on the new population, the fitness value calculation and offspring individual generation steps and the new population generation step are repeated until the preset number of iterations is reached. In other words, the number of evolutions (iterations) is used as the condition for the algorithm to terminate. When the number of iterations exceeds a certain number, the algorithm is terminated and the individual with the smallest fitness value in the new population when the preset number of iterations is reached is taken as the optimal solution of the comprehensive objective function, thereby maximizing the total ore transportation volume and minimizing the total voyage cost.
[0110] Based on the actual scale of operation, data is generated and numerical experiments are conducted to verify the applicability of the voyage optimization method (involving a comprehensive objective function and a non-dominated sorting genetic algorithm) for the routes of dry bulk carriers using new fuels of the present invention.
[0111] This case is based on two key parameters: the time range T of the voyage and the duration of the voyage (or the time span of the voyage) P. The detailed configuration is shown in Table 5. Other basic parameters are generated or selected according to the actual range: the number of ships in the fleet is set to 16; according to the loading capacity of the ore port collected in the historical data, it is averaged in each time period, and the loading capacity Q of the ship is 210,000 tons; the speed set V is the real number set {9, 10, 11, 12, 13}; the cost is the unit fuel consumption collected from the historical data; the unloading time t0 is set to 5 days; the shortest and longest voyage times as well as the mean and variance of the voyage time are all obtained from historical data.
[0112] Table 5
[0113]
[0114] The computation time in the experimental results does not include the time for data processing. After weighing the computation time and the quality of the solution, the parameters of NSGA-II are set as follows: population size is 50; iteration time is 10; crossover rate is 0.9; mutation rate is 0.9; and variation rate is 0.1.
[0115] Then, computational experiments are performed to evaluate the performance of effective inequalities (14) and (15). As shown in Table 6, the computational results under different configurations of T and P are reported. "Time (1)", "Gap (1)", "Time (2)", and "Gap (2)" are the computational time (s) and optimization degree of solving with and without effective inequalities, respectively. By comparison, it can be seen that the good effect of effective inequalities in accelerating the solution process is demonstrated. Due to the addition of effective inequalities, the computational time is reduced from 3600 seconds to only 278.58 seconds.
[0116] Table 6
[0117]
[0118]
[0119] In addition, the effectiveness of the NSGA-II algorithm is also verified. Table 7 lists multiple solutions using the CPLEX algorithm and the NSGA-II algorithm used in the present invention. It can be seen that the use of NSGA-II can provide more Pareto solutions within a reasonable computing time for managers to weigh and choose according to actual needs.
[0120] Table 7
[0121]
[0122] The present invention also relates to a voyage optimization system for a new fuel dry bulk carrier route. The system corresponds to the voyage optimization method for the new fuel dry bulk carrier route, and can be understood as a system for implementing the above method. The system includes a parameter acquisition module, an objective function establishment module, an initial population generation module, a fitness value calculation and offspring individual generation module, a new population generation module and an optimal solution calculation module connected in sequence. Specifically,
[0123] The parameter acquisition module acquires the ship parameters and navigation parameters of the new fuel dry bulk carrier, wherein the ship parameters include the maximum loading capacity of the ship, the fleet size, the minimum number of voyages, and the cumulative loading capacity of the ship at the loading port; the navigation parameters include the starting time of the voyage, the duration of the voyage, and the cost of the voyage;
[0124] The objective function establishment module constructs decision variables based on the judgment result of whether the ship starts sailing at the starting sailing time of the voyage and the sailing duration, and establishes a first objective function according to the maximum loading capacity of the ship and the decision variables under multiple constraint conditions established by the value range of the decision variables, the limit range of the fleet size, the minimum number of voyages, and the limit range of the cumulative loading volume of the ship at the loading port as constraints, and establishes a second objective function according to the voyage cost determined by the starting sailing time and the sailing duration of the voyage and the decision variables, and then merges the first objective function and the second objective function based on a preset weight coefficient and a weighted sum method to obtain a comprehensive objective function;
[0125] The initial population generation module divides the total sailing time of the ship into multiple time periods in chronological order to generate an initial population, and regards each time period in the initial population as an individual;
[0126] The fitness value calculation and offspring individual generation module calculates the fitness value of each individual in the initial population according to the first objective function value and its weight coefficient, and the second objective function value and its weight coefficient, and uses the genetic operator in the non-dominated sorting genetic algorithm to perform a neighborhood search evolution operation on multiple individuals in the initial population to generate multiple new offspring individuals, and calculates the fitness value of each new offspring individual;
[0127] The new population generation module merges the new offspring individuals with the parent individuals to generate a temporary population, and compares the number of individuals in the temporary population with a preset number threshold. If the number of individuals in the temporary population is greater than the preset number threshold, the fitness value of each individual in the temporary population is sorted according to size, and the individuals exceeding the number threshold in the sorted temporary population are deleted in order from small to large to obtain a new population;
[0128] The optimal solution calculation module is based on the new population and repeats the fitness value calculation and the work of the offspring individual generation module and the new population generation module until a preset number of iterations is reached. The individual with the smallest fitness value in the new population when the preset number of iterations is reached is taken as the optimal solution of the comprehensive objective function, thereby maximizing the total dry bulk cargo transshipment volume and minimizing the total voyage cost.
[0129] Preferably, in the fitness value calculation and offspring individual generation module, the genetic operators include a selection operator, a new blood operator, a crossover operator, an insertion operator and a mutation operator; using the genetic operators in the non-dominated sorting genetic algorithm to perform neighborhood search evolution operations on multiple individuals in the initial population to generate multiple new offspring individuals specifically includes:
[0130] S1: The selection operator selects multiple individuals as parent individuals from the initial population according to the fitness value and adopts the roulette wheel selection method;
[0131] S2: The new blood operator randomly generates multiple new individuals and adds them to the parent individuals;
[0132] S3: Determine whether to perform a crossover operation based on the randomly generated probability value and the preset crossover rate. If yes, the crossover operator randomly selects two individuals from the parent individuals, and then randomly selects two time units with relative positions from the two individuals, and exchanges the codes of the two time units to generate two offspring individuals; if no, the number of parent individuals remains unchanged, and all individuals in the parent individuals are used as offspring individuals;
[0133] S4: Repeat step S3 until the preset number of iterations is reached to obtain multiple offspring individuals;
[0134] S5: The insertion operator performs an insertion operation on each offspring individual to obtain multiple inserted offspring individuals;
[0135] S6: The mutation operator performs mutation operations on each offspring individual after insertion to obtain multiple new offspring individuals.
[0136] Preferably, the parameter acquisition module also acquires port parameters, which include the initial inventory of the loading port and the loading capacity of the loading port at a certain moment.
[0137] Preferably, in the objective function establishment module, the multiple constraints include a first constraint, a second constraint, a third constraint, a fourth constraint, a fifth constraint, a sixth constraint, a seventh constraint and an eighth constraint; the first constraint is constructed according to a decision variable and a preset numerical threshold, the second constraint is constructed according to the decision variable and the fleet size, the third constraint is constructed according to the decision variable and the minimum number of voyages, the fourth constraint is constructed according to the cumulative loading of the ship at the loading port and the maximum loading of the ship, the fifth constraint is constructed according to the cumulative loading of the ship at the loading port, the initial inventory of the loading port, the loading capacity of the loading port at a certain time, the maximum loading of the ship and the decision variables, the sixth constraint is constructed according to the cumulative loading of the ship at the loading port, the loading capacity of the loading port at a certain time, the maximum loading of the ship and the decision variables, the seventh constraint is constructed according to the initial inventory of the loading port, the maximum loading of the ship and the decision variables, and the eighth constraint is constructed according to the cumulative loading of the ship at the loading port, the maximum loading of the ship and the decision variables.
[0138] Preferably, in the initial population generation module, each time period is further divided into a plurality of time units in chronological order, and each time unit is encoded to obtain the encoded time units.
[0139] The present invention provides an objective and scientific voyage optimization method and system for new fuel dry bulk carrier routes, with the goals of maximizing the total dry bulk cargo transshipment volume and minimizing the total voyage cost, and taking the value range of decision variables, the limit range of fleet size, the minimum number of voyages, the limit range of the cumulative loading volume of the ship at the loading port, etc. as constraints, establishes the first and second objective functions, and then constructs a comprehensive objective function, and designs a customized non-dominated sorting genetic algorithm to obtain the optimal solution of the comprehensive objective function, so as to maximize the total dry bulk cargo transshipment volume while minimizing the total voyage cost, which can effectively improve operational efficiency and reduce costs.
[0140] 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 voyage optimization method for a new fuel dry bulk carrier route, characterized in that: The following steps are involved: Parameter acquisition step: acquiring ship parameters and navigation parameters of the new fuel dry bulk carrier, wherein the ship parameters include the maximum loading capacity of the ship, the fleet size, the minimum number of voyages, and the cumulative loading capacity of the ship at the loading port; the navigation parameters include the starting time of the voyage, the duration of the voyage, and the cost of the voyage; Objective function establishment step: constructing decision variables based on the judgment result of whether the ship starts sailing at the starting sailing time of the voyage and the sailing duration, and establishing a first objective function according to the maximum loading capacity of the ship and the decision variables under multiple constraints established by the value range of the decision variables, the limit range of the fleet size, the minimum number of voyages, and the limit range of the cumulative loading volume of the ship at the loading port as constraints, and establishing a second objective function according to the voyage cost determined by the starting sailing time and the sailing duration of the voyage and the decision variables, and then merging the first objective function and the second objective function based on the preset weight coefficient and the weighted sum method to obtain a comprehensive objective function; Initial population generation step: the total sailing time of the ship is divided into multiple time periods in chronological order to generate the initial population, and each time period in the initial population is regarded as an individual; Fitness value calculation and offspring individual generation step: the fitness value of each individual in the initial population is calculated according to the first objective function value and its weight coefficient, and the second objective function value and its weight coefficient, and the genetic operator in the non-dominated sorting genetic algorithm is used to perform neighborhood search evolution operation on multiple individuals in the initial population to generate multiple new offspring individuals, and the fitness value of each new offspring individual is calculated; New population generation step: merge the new offspring individuals with the parent individuals to generate a temporary population, and compare the number of individuals in the temporary population with the preset number threshold. If the number of individuals in the temporary population is greater than the preset number threshold, sort the fitness value of each individual in the temporary population by size, and delete the individuals that exceed the number threshold in the sorted temporary population from small to large to obtain a new population; Optimal solution calculation steps: Based on the new population, the fitness value calculation and offspring individual generation steps and the new population generation steps are repeated until the preset number of iterations is reached. The individual with the smallest fitness value in the new population when the preset number of iterations is reached is taken as the optimal solution of the comprehensive objective function, thereby maximizing the total dry bulk cargo transshipment volume and minimizing the total voyage cost.
2. The voyage optimization method for new fuel dry bulk carrier routes according to claim 1, characterized in that: In the fitness value calculation and offspring individual generation steps, the genetic operators include a selection operator, a new blood operator, a crossover operator, an insertion operator and a mutation operator; using the genetic operators in the non-dominated sorting genetic algorithm to perform neighborhood search evolution operations on multiple individuals in the initial population to generate multiple new offspring individuals specifically includes: S1: The selection operator selects multiple individuals as parent individuals from the initial population according to the fitness value and adopts the roulette wheel selection method; S2: The new blood operator randomly generates multiple new individuals and adds them to the parent individuals; S3: Determine whether to perform a crossover operation based on the randomly generated probability value and the preset crossover rate. If yes, the crossover operator randomly selects two individuals from the parent individuals, and then randomly selects two time units with relative positions from the two individuals, and exchanges the codes of the two time units to generate two offspring individuals; if no, the number of parent individuals remains unchanged, and all individuals in the parent individuals are used as offspring individuals; S4: Repeat step S3 until the preset number of iterations is reached to obtain multiple offspring individuals; S5: The insertion operator performs an insertion operation on each offspring individual to obtain multiple inserted offspring individuals; S6: The mutation operator performs mutation operations on each inserted offspring individual to obtain multiple new offspring individuals.
3. The voyage optimization method for new fuel dry bulk carrier routes according to claim 1 or 2, characterized in that: In the parameter acquisition step, port parameters are also acquired, and the port parameters include the initial inventory of the loading port and the loading capacity of the loading port at a certain moment.
4. The voyage optimization method for new fuel dry bulk carrier routes according to claim 3, characterized in that: In the objective function establishment step, the multiple constraints include a first constraint, a second constraint, a third constraint and a fourth constraint, the first constraint is constructed based on the decision variable and a preset numerical threshold, the second constraint is constructed based on the decision variable and the fleet size, the third constraint is constructed based on the decision variable and the minimum number of voyages, and the fourth constraint is constructed based on the cumulative loading volume of the ship at the loading port and the maximum loading volume of the ship.
5. The voyage optimization method for new fuel dry bulk carrier routes according to claim 4, characterized in that: In the objective function establishment step, the multiple constraints also include a fifth constraint, a sixth constraint, a seventh constraint and an eighth constraint. The fifth constraint is constructed based on the cumulative loading volume of the ship at the loading port, the initial inventory of the loading port, the loading capacity of the loading port at a certain moment, the maximum loading volume of the ship and the decision variables. The sixth constraint is constructed based on the cumulative loading volume of the ship at the loading port, the loading capacity of the loading port at a certain moment, the maximum loading volume of the ship and the decision variables. The seventh constraint is constructed based on the initial inventory of the loading port, the maximum loading volume of the ship and the decision variables. The eighth constraint is constructed based on the cumulative loading volume of the ship at the loading port, the maximum loading volume of the ship and the decision variables.
6. The voyage optimization method for new fuel dry bulk carrier routes according to claim 1, characterized in that: In the initial population generation step, each time period is divided into a plurality of time units in chronological order, and each time unit is encoded to obtain the encoded time units.
7. A voyage optimization system for new fuel dry bulk carrier routes, characterized in that: It includes a parameter acquisition module, an objective function establishment module, an initial population generation module, a fitness value calculation and offspring individual generation module, a new population generation module and an optimal solution calculation module, which are connected in sequence. The parameter acquisition module acquires the ship parameters and navigation parameters of the new fuel dry bulk carrier, wherein the ship parameters include the maximum loading capacity of the ship, the fleet size, the minimum number of voyages, and the cumulative loading capacity of the ship at the loading port; the navigation parameters include the starting time of the voyage, the duration of the voyage, and the cost of the voyage; The objective function establishment module constructs decision variables based on the judgment result of whether the ship starts sailing at the starting sailing time of the voyage and the sailing duration, and establishes a first objective function according to the maximum loading capacity of the ship and the decision variables under multiple constraint conditions established by the value range of the decision variables, the limit range of the fleet size, the minimum number of voyages, and the limit range of the cumulative loading volume of the ship at the loading port as constraints, and establishes a second objective function according to the voyage cost determined by the starting sailing time and the sailing duration of the voyage and the decision variables, and then merges the first objective function and the second objective function based on a preset weight coefficient and a weighted sum method to obtain a comprehensive objective function; The initial population generation module divides the total sailing time of the ship into multiple time periods in chronological order to generate an initial population, and regards each time period in the initial population as an individual; The fitness value calculation and offspring individual generation module calculates the fitness value of each individual in the initial population according to the first objective function value and its weight coefficient, and the second objective function value and its weight coefficient, and uses the genetic operator in the non-dominated sorting genetic algorithm to perform a neighborhood search evolution operation on multiple individuals in the initial population to generate multiple new offspring individuals, and calculates the fitness value of each new offspring individual; The new population generation module merges the new offspring individuals with the parent individuals to generate a temporary population, and compares the number of individuals in the temporary population with a preset number threshold. If the number of individuals in the temporary population is greater than the preset number threshold, the fitness value of each individual in the temporary population is sorted according to size, and the individuals exceeding the number threshold in the sorted temporary population are deleted in order from small to large to obtain a new population; The optimal solution calculation module is based on the new population and repeats the fitness value calculation and the work of the offspring individual generation module and the new population generation module until a preset number of iterations is reached. The individual with the smallest fitness value in the new population when the preset number of iterations is reached is taken as the optimal solution of the comprehensive objective function, thereby maximizing the total dry bulk cargo transshipment volume and minimizing the total voyage cost.
8. The voyage optimization system for new fuel dry bulk carrier routes according to claim 7, characterized in that: In the fitness value calculation and offspring individual generation module, the genetic operators include a selection operator, a new blood operator, a crossover operator, an insertion operator and a mutation operator; using the genetic operators in the non-dominated sorting genetic algorithm to perform neighborhood search evolution operations on multiple individuals in the initial population to generate multiple new offspring individuals specifically includes: S1: The selection operator selects multiple individuals as parent individuals from the initial population according to the fitness value and adopts the roulette wheel selection method; S2: The new blood operator randomly generates multiple new individuals and adds them to the parent individuals; S3: Determine whether to perform a crossover operation based on the randomly generated probability value and the preset crossover rate. If yes, the crossover operator randomly selects two individuals from the parent individuals, and then randomly selects two time units with relative positions from the two individuals, and exchanges the codes of the two time units to generate two offspring individuals; if no, the number of parent individuals remains unchanged, and all individuals in the parent individuals are used as offspring individuals; S4: Repeat step S3 until the preset number of iterations is reached to obtain multiple offspring individuals; S5: The insertion operator performs an insertion operation on each offspring individual to obtain multiple inserted offspring individuals; S6: The mutation operator performs mutation operations on each offspring individual after insertion to obtain multiple new offspring individuals.
9. The voyage optimization system for new fuel dry bulk carrier routes according to claim 7 or 8, characterized in that: In the parameter acquisition module, port parameters are also acquired, and the port parameters include the initial inventory of the loading port and the loading capacity of the loading port at a certain moment.
10. The voyage optimization system for new fuel dry bulk carrier routes according to claim 9, characterized in that: In the objective function establishment module, the multiple constraints include a first constraint, a second constraint, a third constraint, a fourth constraint, a fifth constraint, a sixth constraint, a seventh constraint and an eighth constraint; the first constraint is constructed according to a decision variable and a preset numerical threshold, the second constraint is constructed according to a decision variable and a fleet size, the third constraint is constructed according to a decision variable and a minimum number of voyages, the fourth constraint is constructed according to a cumulative loading of the ship at the loading port and a maximum loading of the ship, the fifth constraint is constructed according to a cumulative loading of the ship at the loading port, an initial inventory of the loading port, a loading capacity of the loading port at a certain moment, a maximum loading of the ship and decision variables, the sixth constraint is constructed according to a cumulative loading of the ship at the loading port, a loading capacity of the loading port at a certain moment, a maximum loading of the ship and decision variables, the seventh constraint is constructed according to an initial inventory of the loading port, a maximum loading of the ship and decision variables, and the eighth constraint is constructed according to a cumulative loading of the ship at the loading port, a maximum loading of the ship and decision variables.
Citation Information
Patent Citations
Flight route optimization method based on multi-target adaptive weight evolution algorithm
CN113191571A
Shipborne refrigerated container integrated system
CN116424539A
System for optimizing transportation scheduling
SG160982A1
Cited By
Multi-ship route dynamic optimization method and device
CN119761958A
A Route Dynamic Optimization Method and Device for Multiple Ships
CN119761958B
