Method and apparatus for coordinated allocation of continuous berths and time-varying quay cranes at a terminal

Through the three-stage cultural gene algorithm, the continuous berth and time-varying shore bridge allocation of the dock is optimized, and the efficiency control and coordination problems in the dock are solved, efficient berth and shore bridge allocation is achieved, reducing the ship's offset distance and improving the dock Operational efficiency.

CN117391329BActive Publication Date: 2024-12-03GUANGZHOU INSTITUTE OF TECHNOLOY XIDIAN UNIVERSITY
View PDF 2 Cites 0 Cited by

Patent Information

Application Number
CN202311102069.0
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2023-08-30
Publication Date
2024-12-03
Estimated Expiration
2043-08-30

AI Technical Summary

Technical Problem

The prior art is difficult to achieve optimal efficiency control and high coordination in the dock, especially in the integration of continuous berth allocation and time-varying shore bridge allocation, which cannot effectively deal with ship offset distance and economic costs, and solve large-scale problems. Slow speed or infeasible docking schemes.

Method used

Using three-stage cultural gene algorithm (ETMA), through sorting encoding, initial de-optimization, clustering and evolutionary iteration, combined with local search and elite retention strategies, optimize berth and shore bridge allocation, maximize shore bridge utilization and reduce ship bias Move the distance.

Benefits of technology

It realizes the rapid finding of convergent solutions in large-scale problems, improves the operating efficiency of docks, reduces the offset distance of ships, avoids meaningless searches, and ensures the excellence and efficiency of solutions.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN117391329B_ABST
    Figure CN117391329B_ABST
Patent Text Reader

Abstract

The present invention discloses a method and device for collaborative allocation of continuous berths and time-varying quay cranes for a wharf, the method comprising: sorting and encoding each ship individual according to the ship serial number; obtaining a rough feasible initial solution of the first stage; allocating quay cranes to each ship in a time-varying manner, and updating the ship's starting berthing time and working time at the same time; clustering the ships to generate an initial population; evolving and iterating the initial population based on a fitness function; performing a local search operation on the optimal individual of each generation of the population using a preset local search operator; and obtaining the optimal individual that is the optimal solution of the preset model. The present invention provides an innovative method and device for time-varying allocation of ship berths and quay cranes, which significantly improves the operating efficiency of port terminals and achieves the purpose of maximizing the utilization rate of quay cranes and improving the efficiency of terminal operations. The invention has broad application prospects in optimizing the problem of joint resource allocation.
Need to check novelty before this filing date? Find Prior Art

Description

[Technical field]

[0001] The present invention relates to the technical field of wharf berth allocation and quay crane allocation, and in particular to a method and device for collaboratively allocating continuous berths and time-varying quay cranes at a wharf. [Background technology]

[0002] In terminal operations, the rational allocation of two scarce resources, berths and quay cranes (QC), is very important. Berth allocation problem (BAP) is one of the main problems in terminal operations. According to the type of terminal, it can be divided into discrete berth allocation problem (DBAP) and continuous berth allocation problem (CBAP). For DBAP, the terminal is divided into multiple berths, each berth can only serve one ship at a time, and the berth and berthing order of the ship need to be determined. For CBAP, the terminal is a whole, and the ship can be berthed at any position of the terminal, and the berthing position and berthing time of the ship need to be determined.

[0003] At the same time, as one of the key equipment for terminal operation, quay cranes are professional equipment specially used for container terminals to load and unload container ships. The number of quay cranes assigned to ships determines the working efficiency of the ships. The quay crane assignment problem (QCAP) can be divided into time-invariant quay crane assignment and time-varying quay crane assignment. In the time-invariant quay crane assignment, the number of quay cranes assigned to the ship does not change during the working period until the ship leaves the port. In the time-varying quay crane assignment, the number of quay cranes assigned is allowed to change during the berthing time of the ship. This strategy can effectively improve the utilization rate of quay cranes and arrange the berthing time for the ship more reasonably. In BAP, the working time of the ship depends on the number of quay cranes assigned to the ship and is related to the berthing position. Therefore, how to reasonably allocate berths and quay cranes to ships is one of the important issues for efficient control and high coordination of the terminal.

[0004] Among them, the integrated berth allocation and quay crane assignment problem (BACAP) is a combination of BAP and QCAP. BACAP can be divided into continuous and discrete types according to the type of BAP, and quay crane assignment can also be divided into time-invariant and time-varying. At present, most of the research on continuous BACAP is concentrated, but most of the research chooses to divide the continuous berth into many small berth segments, or convert the continuous problem into a discrete type before solving it. Research has shown that this treatment can effectively improve the calculation speed, but for the continuous berth assignment problem, when the berthing position of the ship deviates from the expected position, it often produces certain economic costs and unfeasible berthing solutions, such as some larger ships cannot berth in shallow waters. However, the research on continuous BACAP without discretization is rarely combined with the time-varying quay crane assignment strategy. Moreover, most of the research takes minimizing the ship's stay time in the port as the ultimate goal to establish various mathematical models, and rarely considers the ship's offset distance, so it is impossible to achieve optimal efficiency control and high coordination of the terminal. In addition, the current solution methods used in BACAP research are mainly divided into two types: exact algorithms and approximate algorithms. Exact algorithms can provide exact solutions to BACAP, but due to their slow solution speed, they often cannot handle large-scale BACAP. Approximate algorithms can give an approximate optimal solution in a short time, but this is not wise for minimizing the offset distance of the target berth, and some unnecessary offsets are often generated in the approximate optimal solution. Memetic Algorithm (MA) is a population-based heuristic search algorithm that combines the characteristics of global search based on a set of candidate populations and local heuristic optimization based on a single candidate individual. Compared with the above exact methods and approximate methods, MA is more suitable for solving large-scale BACAP problems with multiple sets of constraints and joint optimization decision variables.

[0005] In view of this, it is necessary to provide a method and device for coordinated allocation of continuous berths and time-varying quay cranes for a wharf to overcome the above-mentioned defects. [Summary of the invention]

[0006] The purpose of the present invention is to provide a method and device for the coordinated allocation of continuous berths and time-varying quay cranes at a terminal, aiming to solve the problem that the current mathematical model is unable to perform optimal efficiency control and high coordination on the terminal, and to find an excellent solution to the integrated problem of continuous berth allocation and time-varying quay crane allocation.

[0007] In order to achieve the above object, the present invention provides a method for collaboratively allocating continuous berths and time-varying quay cranes for a wharf, comprising:

[0008] Step S10: sorting and coding each ship according to the ship serial number to indicate the berthing order of the ships;

[0009] Step S20: obtaining a rough feasible initial solution of the first stage according to the berthing sequence of the ships, the initial solution including the berth allocated to each ship, the starting berthing time and the working time;

[0010] Step S30: using the initial solution obtained in step S20 as input, allocating a quay crane to each ship in a time-varying manner according to the berth allocation situation, updating the ship's starting berthing time and working time, and outputting a second-stage solution;

[0011] Step S40: further improving the solution obtained in step S30, optimizing the berth of the ship, and outputting a final solution;

[0012] Step S50: clustering the ships according to their target berths and estimated arrival times, and then generating an initial population based on the clustering results;

[0013] Step S60: performing evolution iteration on the initial population based on a preset fitness function;

[0014] Step S70: performing a local search operation on the best individual of each generation of population using a preset local search operator to improve and optimize the best individual of each generation of population;

[0015] Step S80: Apply the elite retention strategy to retain the best preset individuals in each generation to the next generation. When the preset iteration termination condition is reached, the optimal individual that is the optimal solution of the preset model is obtained.

[0016] In a preferred embodiment, the step S20 includes:

[0017] The working hours of the corresponding ships are calculated according to the allocated preset number of quay cranes, and then the berths and starting berthing time are allocated to each ship in turn according to the berthing order of the ships;

[0018] Specifically, it is determined whether ship i overlaps with the moored ships in time and space. If so, feasible berthing segments are searched in chronological order.

[0019] In a preferred embodiment, the step S30 includes:

[0020] Step S31: Divide the time periods according to the initial solution obtained in step S20 so that the number of ships at the dock does not change in each time period, and obtain the starting timetable time_period of each time period;

[0021] Step S32: at the beginning of each time period, a quay crane is allocated to each ship according to the proportion of the remaining workload of the ships at the dock, and the number of quay cranes must also meet the limit;

[0022] Step S33: If the ship i can complete the work within the next time period, the berthing time of the ship i is shortened and the departure time of the ship i is updated;

[0023] Step S34: searching for a ship h whose berth overlaps with that of the ship i, and accordingly advancing the berthing time of the ship h to shorten the waiting berthing time of the ship h, and updating the start schedule time_period.

[0024] In a preferred embodiment, the step S40 includes:

[0025] Detect the berthing position of each ship and move it horizontally to get it as close to its target berth as possible.

[0026] In a preferred embodiment, the step S50 includes: step S51: using a preset clustering operator to divide the set of ships into K categories, and sorting the ships according to their estimated arrival times;

[0027] Step S52: on the basis of the ordering by estimated arrival time, the order of the ships in the same cluster after clustering is randomly sorted to obtain different individuals;

[0028] Step S53: Generate an initial population based on the result of step S52.

[0029] In a preferred embodiment, the step S60 includes:

[0030] Step S61: selecting a parent generation using a roulette algorithm according to the value of the fitness function;

[0031] Step S62: randomly selecting a cluster from the clustering results by using a preset crossover operator, then randomly selecting two crossover points in the cluster, and sequentially crossovering the two selected parent generations to be crossovered;

[0032] Step S63: randomly selecting two clusters from the clustering results through a preset mutation operator, and randomly selecting two points in the two clusters for exchange to complete the mutation of the individual.

[0033] In a preferred embodiment, in step S70, the following operations are performed on the ship in descending order of the objective function value:

[0034] Step S71: Check whether the ship can directly berth at the target berth at the estimated arrival time. If there is an overlap with other ships, search for a feasible berth section in the current time period;

[0035] Step S72: while keeping the berthing time unchanged, check whether the idle quay cranes in the current period are sufficient for the ship to complete the work; if not, try to extend the berthing time of the ship, otherwise go to step S73;

[0036] Step S73: traverse from the estimated arrival time to the allocated parking time, and check whether there is a parking time for the berth segment that can be parked. If so, repeat the above steps S71-S72, if not, stop.

[0037] In a preferred embodiment, the steps are also included:

[0038] Determine whether the number of iterations is a multiple of G. If the result is yes, re-cluster; if the result is no, retain the elite individuals in the population; where G is a preset positive integer and G≥2.

[0039] In a preferred embodiment, in step S60, the fitness function is set to the inverse of the objective function; wherein the objective function is: V is the set of ships, V={1,2,…,|V|}, a i is the estimated arrival time of ship i, ts i is the start berthing time of ship i, b i is the target berth position of ship i, p i is the berthing position of ship i, α, β are the weight values ​​of the preset model target.

[0040] The present invention also provides a coordinated allocation device for continuous berths and time-varying quay cranes at a wharf, comprising:

[0041] The ship coding module is used to sort and code each ship according to its serial number to indicate the berthing order of the ship;

[0042] The first stage decoding module is used to obtain a rough feasible initial solution of the first stage according to the berthing sequence of the ships, and the initial solution includes the berth allocated to each ship and the starting berthing time;

[0043] The second stage decoding module is used to use the initial solution obtained by the first stage decoding module as input, allocate a quay crane to each ship in a time-varying manner according to the berth allocation situation, update the ship's starting berthing time and working time, and output the second stage solution;

[0044] The third stage decoding module is used to improve the solution obtained by the second stage decoding module, optimize the berth of the ship, and output the final solution;

[0045] An initial population generation module is used to cluster the ships according to their target berths and estimated arrival times, and then generate an initial population based on the clustering results;

[0046] A population evolution module, used for iterating the evolution of the initial population based on the fitness function;

[0047] The individual optimization module is used to perform local search operations on the best individuals of each generation of population based on a preset local search operator, and to improve and optimize the best individuals of each generation of population;

[0048] The elite retention module is used to apply the elite retention strategy to retain the best preset individuals in each generation to the next generation. When the preset iteration termination condition is reached, the optimal individual that is the optimal solution of the preset model is obtained.

[0049] The method and device for collaborative allocation of continuous berths and time-varying quay cranes for a wharf provided by the present invention, by executing steps S10-S40, while allocating berths to ships, also allocates quay cranes to ships in a time-varying manner, thereby achieving the purpose of maximizing the utilization rate of quay cranes and improving the efficiency of wharf operations. Then, by executing steps S50-S80, the ships are clustered according to the spatiotemporal data of the estimated arrival time of the ships and the target berths, and diverse and potential individuals are generated according to the clustering results. Then, the population is subjected to crossover and mutation operators designed based on the clustering results, and the target-guided local search operator is executed, and finally, an elite retention strategy is adopted to retain excellent individuals and generate offspring. The population initialization method and evolution operator based on the clustering results effectively narrow the search range of excellent solutions, avoid meaningless searches, and improve search efficiency, so that the algorithm can still quickly find a convergent solution when processing large-scale instances. At the same time, the target-guided local search operator is further optimized on the basis of finding an excellent solution to ensure the excellence of the solution.

Brief Description of the Drawings

[0050] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for use in the embodiments are briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present invention and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without creative work.

[0051] Figure 1 An algorithm flow chart of the method for collaboratively allocating continuous berths and time-varying quay cranes for a wharf provided by the present invention;

[0052] Figure 2 A schematic diagram of a crossover operator in step S50 of the method for collaboratively allocating continuous berths and time-varying quay cranes for a wharf provided by the present invention;

[0053] Figure 3A schematic diagram of a mutation operator in step S50 of the method for collaboratively allocating continuous berths and time-varying quay cranes for a wharf provided by the present invention;

[0054] Figure 4 A flow chart of a local search operator in step S60 of the method for collaboratively allocating continuous berths and time-varying quay cranes for a wharf provided by the present invention;

[0055] Figure 5 It is a schematic diagram of candidate solutions on the BACAP problem model;

[0056] Figure 6 It is a schematic diagram of the convergence speed of each algorithm in a specific embodiment;

[0057] Figure 7 Schematic diagram of the visualization results of the optimal solution obtained by the ETMA algorithm. [Specific implementation method]

[0058] In order to make the purpose, technical solution and beneficial technical effect of the present invention clearer, the present invention is further described in detail below in conjunction with the accompanying drawings and specific implementation methods. It should be understood that the specific implementation methods described in this specification are only for explaining the present invention, not for limiting the present invention.

[0059] It should also be understood that the terms used in this specification of the present invention are only for the purpose of describing specific embodiments and are not intended to limit the present invention. As used in the specification of the present invention and the appended claims, unless the context clearly indicates otherwise, the singular forms "a", "an" and "the" are intended to include plural forms.

[0060] It should be further understood that the term "and / or" used in the present description and the appended claims refers to any and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0061] In an embodiment of the present invention, a method for collaborative allocation of continuous berths and time-varying quay cranes for a terminal is provided for efficiency control and high coordination of the terminal, a mathematical model of the collaborative optimization problem (BACAP) of continuous berth allocation and time-varying allocation of quay cranes with collaborative objectives of ship offset distance and waiting time is established, and an enhanced three-stage memetic algorithm (ETMA) is designed based on the problem model, so that continuous berths can be truly processed. Different from the existing research on discrete processing of berths, the present invention pays special attention to the continuous BACAP problem that has not been discretized. The additional cost and unfeasible berthing schemes incurred when the berthing position of the ship deviates from the expected position are considered, especially the situation that large ships may not be able to berth in shallow waters. Therefore, the offset distance of the ship is minimized when allocating berths to ships; the present invention combines continuous BACAP with time-varying quay crane allocation strategy to ensure that the ship loading and unloading operations are optimally scheduled, thereby improving the overall port efficiency, maximizing the utilization rate of quay cranes, and improving the efficiency of terminal operations; unlike the past focus on minimizing the time that ships stay in port, the present invention introduces the ship offset distance as an additional goal, realizing the comprehensive control of the optimal efficiency and coordination of the port. In summary, the collaborative optimization method (BACAP) of berth allocation and quay crane allocation proposed in the present invention has the advantages of comprehensive consideration of offset costs, true continuous berth allocation, flexible quay crane allocation, and large-scale problem handling, providing a more intelligent and efficient solution for port operations.

[0062] Combination Figure 1 As shown, the method for collaboratively allocating continuous berths and time-varying quay cranes for a terminal includes steps S10-S80.

[0063] Before step S10, a fitness function may be defined. Since the BACAP problem is a minimization optimization problem, the fitness function is set to the inverse of the objective function. The objective function may be set to: V is the set of ships, V={1,2,…,|V|}, a i is the estimated arrival time of ship i, ts i is the start berthing time of ship i, b i is the target berth position of ship i, p i is the berthing position of ship i, α, β are the weight values ​​of the preset model target.

[0064] Step S10: sort and encode each ship according to its serial number to indicate the berthing order of the ships.

[0065] The ETMA algorithm provided by the present invention uses sorting coding, and uses the serial number of the ship for coding to indicate the berthing order of the ship. i The chromosomes of can be represented as follows: i ={1,2,…,|V|}.

[0066] Step S20: A rough feasible initial solution of the first stage is obtained according to the berthing sequence of the ships. The initial solution includes the berth allocated to each ship, the starting berthing time and the working time.

[0067] Here, step S20 includes: calculating the working time of the corresponding ship according to the assigned preset number of quay cranes, and then allocating berths and starting berthing time for each ship in turn according to the berthing order of the ship (i.e., the chromosome information of this genetic algorithm). It should be noted that, in order to facilitate calculation and only consider shortening the berthing time in the second stage of decoding, we temporarily allocate one quay crane to each ship.

[0068] Specifically, it can be determined first whether ship i can berth at the target berth at the estimated arrival time, that is, whether ship i overlaps with the moored ships in time and space. If there is overlap, feasible berth segments are searched in chronological order.

[0069] Step S30: Using the initial solution obtained in step S20 as input, allocate a quay crane to each ship in a time-varying manner according to the berth allocation situation, update the ship's starting berthing time and working time, and output the solution for the second stage.

[0070] Here, step S30 includes steps S31 - S34 .

[0071] Step S31: Divide the time periods according to the initial solution obtained in step S20 so that the number of ships at the dock does not change within each time period, and obtain the starting schedule time_period of each time period.

[0072] Step S32: At the beginning of each time period (i.e., at the time when each ship enters or leaves the port), a quay crane is allocated to each ship according to the proportion of the remaining workload of the ships at the dock, and the quay crane quantity range limit must be met at the same time.

[0073] Step S33: If the ship i can complete the work within the next time period, the berthing time of the ship i is shortened and the departure time of the ship i is updated.

[0074] Step S34: searching for a ship h whose berth overlaps with that of the ship i, and accordingly advancing the berthing time of the ship h to shorten the waiting berthing time of the ship h, and updating the start schedule time_period.

[0075] Step S40: further improve the solution obtained in step S30, optimize the berth of the ship, and output the final solution, that is, the following steps: detect the berth position of each ship, and move it horizontally to make it as close to its target berth as possible.

[0076] Therefore, the first and second stages of the three-stage decoding method allocate berths to ships and allocate quay cranes to ships in a time-varying manner, thereby maximizing the utilization of quay cranes and improving the efficiency of terminal operations; the third stage further optimizes the target berth offset distance of the ship, truly handles continuous berths, rather than pseudo-continuous berths (actually discretized) as in other patents and documents, thereby minimizing the offset distance of the ship. Therefore, by executing steps S10-S40, while allocating berths to ships, quay cranes are also allocated to ships in a time-varying manner, thereby maximizing the utilization of quay cranes and improving the efficiency of terminal operations.

[0077] It can be understood that the three-stage decoding is to generate a feasible initial solution, allocate the quay crane in a time-varying manner, and optimize the berth position. In some embodiments, the three-stage decoding can be replaced by a two-stage decoding, that is, first generate a feasible initial solution, and then allocate the quay crane in a time-varying manner, but this will produce a poor solution, thereby increasing the search burden and failing to minimize the offset distance of the ship.

[0078] Step S50: clustering the ships according to their target berths and estimated arrival times, and then generating an initial population based on the clustering results.

[0079] In order to narrow the algorithm search scope, the present invention clusters the ships according to their target berths and estimated arrival times. Ships in the same cluster have a high correlation in target berths or estimated arrival times, and then generate an initial population based on the clustering results.

[0080] Specifically, step S50 includes steps S51-S53.

[0081] Step S51: using a preset clustering operator to divide the ship set into K categories, and sorting the ships according to their estimated arrival times. For example, K=4.

[0082] Step S52: On the basis of sorting by estimated arrival time, the order of the ships in the same cluster after clustering is randomized to obtain different individuals.

[0083] Step S53: Generate an initial population based on the result of step S52.

[0084] Step S60: performing evolution iteration on the initial population based on a preset fitness function.

[0085] It should be noted that the evolution operator is applied to perform evolution iteration on the population. The present invention adopts the roulette strategy to select the parent generation. Since the greater the fitness value of the individual in the roulette strategy algorithm, the greater the probability of its selection. At the same time, since the fitness function is the inverse of the objective function, the probability of the individual being selected is inversely proportional to its objective function value. The recombination operator is an improved sequential crossover operator and a mutation operator based on the initial population generation method.

[0086] Specific, combined Figure 2 and Figure 3 As shown, step S60 includes steps S61-S63.

[0087] Step S61: Select the parent generation using the roulette algorithm according to the value of the fitness function.

[0088] Step S62: Figure 2 As shown, a cluster is randomly selected from the clustering results (different colors in the chromosome represent different clusters) through a preset crossover operator, and then two crossover points are randomly selected in the cluster, and sequential crossover is performed on the two selected parents for crossover.

[0089] Step S63: Figure 3 As shown, two clusters are randomly selected from the clustering results through the preset mutation operator, and two points are randomly selected in the two clusters for exchange to complete the mutation of the individual.

[0090] Step S70: Perform a local search operation on the best individual of each generation of population using a preset local search operator, and improve and optimize the best individual of each generation of population. That is, the solution obtained in step S60 is improved by applying a target-guided local search operator. The present invention performs a local search operation on the best individual of each generation of population, and performs the optimization operation on the ships in the order of the objective function value from large to small. Figure 4 Operation in the flowchart.

[0091] Specific, combined Figure 4 As shown, in step S70, the following operations are performed on the ship in descending order of the objective function values:

[0092] Step S71: Check whether the ship can directly berth at the target berth at the estimated arrival time. If there is an overlap with other ships, search for a feasible berth section in the current time period;

[0093] Step S72: while keeping the berthing time unchanged, check whether the idle quay cranes in the current period are sufficient for the ship to complete the work; if not, try to extend the berthing time of the ship, otherwise go to step S73;

[0094] Step S73: traverse from the estimated arrival time to the allocated parking time, and check whether there is a parking time with a berth segment that can be parked. If yes, repeat the above steps S71-S72, if not, stop. Step S80: Apply the elite retention strategy, retain the best preset part of individuals in each generation to the next generation, and when the preset iteration termination condition is reached, the optimal individual that is the optimal solution of the preset model is obtained.

[0095] Therefore, by executing steps S50-S80, the ships are clustered according to the spatiotemporal data of the estimated arrival time and the target berth, and diverse and potential individuals are generated according to the clustering results. Then, the population is subjected to the crossover and mutation operators designed based on the clustering results, and the target-guided local search operator is executed. Finally, the elite protection strategy is adopted to retain the excellent individuals and generate offspring, so that the ships in the problem instance are clustered according to the estimated arrival time and the target berth. The population initialization method and the evolution operator based on the clustering results effectively narrow the search range of the excellent solution, avoid meaningless search, and improve the search efficiency, so that the algorithm can still quickly find the convergent solution when processing large-scale instances. At the same time, the target-guided local search operator executed on the best chromosome so far is further optimized on the basis of finding the excellent solution to ensure the excellence of the solution.

[0096] Furthermore, in some embodiments, in order to prevent the population from falling into a local optimum, the algorithm periodically calls the diversity enhancement method and performs re-clustering every G generations. Specifically, the steps include: judging whether the number of iterations is a multiple of G, if the result is yes, re-clustering; if the result is no, retaining the elite individuals in the population; wherein G is a preset positive integer, and G≥2.

[0097] It should be noted that, in order to facilitate understanding of the above implementation steps and verify the effectiveness of the present invention, a specific embodiment will be given below, and the algorithm proposed by the present invention is applied to this example to give a final solution, and to assist in explaining the above steps. Therefore, some specific parameters and specific processing processes involved in the following specific embodiment do not represent specific limitations on the steps of the present invention.

[0098] In this specific embodiment, it is assumed that the length of the shoreline is 3000 m, there are 30 quay cranes on the shore, the working efficiency of the quay cranes is 25 boxes / h, and berths and quay cranes need to be allocated for 40 ships arriving at the port in the next 300 hours. The input parameter symbols and settings of the ships are shown in Table I. Among them, 10 different data sets are generated according to this setting, and the algorithm is independently run 20 times on the 10 data sets.

[0099] Table I: Vessel input parameter settings

[0100]

[0101] Where V is the set of ships, V = {1, 2, ..., |V|}; Q is the set of quay cranes, Q = {1, 2, ..., |Q|}; i is the ship number; L is the length of the shoreline, in meters (m); l i is the length of ship i; c i is the cargo loading and unloading volume of ship i; a i is the estimated arrival time of ship i; d i is the estimated departure time of vessel i; ts i is the start berthing time of vessel i; tl i is the departure time of ship i; T is a set of one-hour time periods, T={1,2,…,max(tl i )}; t is the number indicating time; b i is the target berth position of ship i; p i is the berthing position of ship i; q i max is the maximum number of quay cranes allowed to be allocated to ship i; q i min is the minimum number of quay cranes required for ship i to complete loading and unloading tasks; v is the working efficiency of the quay cranes; x ij ,y ij ,r ikt ,r mn is a binary variable, where x ij =1 means that ship i is moored to the left of ship j, otherwise it is 0; ij =1 means that the departure time of ship i is not later than the start berthing time of ship j, otherwise it is 0; r ikt =1 means that the quay crane k is allocated to the ship i at time t, otherwise it is 0; r mn =1 means that quay crane m is working on the left side of quay crane n, otherwise it is 0; k is the number of the quay crane; α, β are the weight values ​​of the mathematical model target.

[0102] Then, the BACAP mathematical model is established to minimize the weighted sum of the ship's waiting time for berthing and the target berth offset distance. According to the optimization goal of the problem, the objective function of the problem instance can be obtained:

[0103]

[0104] The setting of the target value weights α and β needs to consider many factors: if the deviation between the two target values ​​is large, the heuristic algorithm tends to ignore the smaller target and focus on optimizing the other larger target. Therefore, when setting the weights, the two target values ​​should be as close as possible: α = 10, β = 1. i -ts i Indicates the waiting time for parking, |bi -p i |Indicates the target berth offset distance.

[0105] In addition, the BACAP problem instance should also satisfy the following constraints:

[0106]

[0107]

[0108]

[0109]

[0110]

[0111]

[0112]

[0113]

[0114]

[0115]

[0116]

[0117]

[0118]

[0119]

[0120] Among them, constraint (2) stipulates that the berthing start time cannot be earlier than the arrival time, ensuring that the waiting time is not negative; constraints (3)-(4) indicate that the ship's berth should be within the coastline; constraints (5)-(6) are non-overlapping constraints, where constraint (5) ensures that the berthing positions of any two ships do not overlap in space and meet the safety distance limit (that is, the distance between the two ships must be greater than or equal to 10% of the ship length); constraint (6) ensures that any two ships do not overlap in time; constraint (7) requires that the quay cranes cannot cross-operate during operation; constraint (8) specifies the maximum number of quay cranes that can be allocated to each ship, because the distance between adjacent quay cranes must be greater than or equal to 24 meters; constraint (9) requires that the number of quay cranes allocated to each ship must be within a given range; constraint (10) requires that the total number of quay cranes allocated to the ship must meet the loading and unloading workload of the ship; constraint (11) is a constraint on the total number of quay cranes; constraints (12)-(15) are constraints on binary variables.

[0121] Figure 5 A simple schematic diagram of the candidate solution on the BACAP problem model is provided, where the horizontal axis represents the shoreline, the vertical axis represents the time, each solid rectangular block represents the berthing situation of a ship (the horizontal axis in the lower left corner of the rectangle represents the berthing position, the vertical axis represents the berthing time, the width of the rectangle represents the length of the ship, and the height represents the working time), the gray box represents the safe distance of the ship (the front and rear of the ship need a safe distance, which is set to 10% of the length of the ship), the five-pointed star represents the estimated arrival time and target berth of each ship, and the arrow can reflect the size of the ship's offset target value. In addition, the small rectangle in the rectangle representing the ship represents the quay bridge assigned to the ship in each time period.

[0122] Then, the algorithm ETMA proposed in the present invention is used to solve the problem instance. First, the fitness function (the inverse of the objective function) is defined:

[0123]

[0124] Then, the ship's serial number is encoded using the sorting code to indicate the ship's berthing order. Then, according to the coded information, the three-stage decoding method designed by ETMA is used for decoding. The specific operation steps are as follows:

[0125] The first step (decoding stage 1 - generating feasible initial solutions): first assign a quay crane to each ship. Calculate the working hours of the ship based on the number of assigned quay cranes, and then assign a berth and start berthing time to each ship based on the chromosome information (the berthing order of the ship). Specifically, first determine whether ship i can berth at the target berth at the estimated arrival time, that is, determine whether ship i overlaps with the berthed ships in time and space. If there is overlap, search for feasible berthing segments in chronological order.

[0126] Step 2 (Decoding Phase 2 - Time-varying allocation of quay cranes): Using the initial solution obtained in Phase 1 as input, allocate quay cranes to each ship in a time-varying manner according to the berth allocation, and update the ship's start berthing time and working time, and output the solution for Phase 2. The specific operation is as follows: First, divide the time period according to the initial solution obtained in Phase 1, so that the ships on the dock do not change in each time period, and obtain the start schedule time_period of each time period. Then, at the beginning of each time period (i.e., at the time point when each ship enters and leaves the port), allocate quay cranes to them according to the proportion of the remaining workload of the ships on the dock, while meeting the quay crane number range limit. Subsequently, if ship i can complete the work in the next time period, shorten the berthing time of ship i, and update the departure time of ship i, search for ship h that has a berth overlapping with ship i, and advance the berthing time of ship h to shorten the waiting berthing time of ship h, and update the start schedule time_period.

[0127] Step 3 (Decoding Phase 3 - Optimizing Berth Position): Detect the berthing position of each ship and move it horizontally to get it as close to its target berth as possible.

[0128] Then generate the initial population, cluster the ships according to their target berths and estimated arrival times, and ships in the same cluster have a high correlation in target berths or estimated arrival times. Then generate the initial population based on the clustering results. The specific operations are as follows: use the clustering operator to divide the ship set into K categories, sort the ships according to their estimated arrival times, and based on this order, randomly cluster the order of the ships in the cluster to obtain different individuals.

[0129] Next, the evolution operator is applied to the population to start evolution iteration. The fitness function is set to the inverse of the objective function, and the roulette strategy is used to select the parent generation. The population is reorganized and mutated using the improved sequential crossover operator and the mutation operator based on the initial population generation method. Figure 2 As shown in , the crossover operator first randomly selects a cluster from the clustering results (different colors in the chromosome represent different clusters), then randomly selects two crossover points in this cluster, and sequentially crosses the two selected parents. Figure 3 As shown, the mutation operator randomly selects two clusters from the clustering results, and randomly selects two points in the two clusters for exchange to complete the mutation of the individual.

[0130] Then, the target-guided local search operator is applied to improve and optimize the best individuals of each generation of population, and the ships are executed in the order of the objective function value from large to small. Figure 4Operations in the flowchart. Specifically, first check whether the ship can berth directly at the target berth at the estimated arrival time. If there is an overlap with other ships, search for a feasible berth segment in the current period. Keep the berthing time unchanged and check whether the idle quay cranes in this period are sufficient for the ship to complete the work. If not, try to extend the berthing time of the ship. Otherwise, traverse from the estimated arrival time to the allocated berthing time to check whether there is a berthing time for the berthing segment. If so, repeat the above operation, if not, stop.

[0131] And the elite retention strategy is applied to retain the best individuals in each generation to the next generation. At the same time, in order to prevent the population from falling into the local optimum, the algorithm regularly calls the diversity enhancement method and re-clustering every G generations.

[0132] Finally, when the iteration termination condition is reached, the optimal individual obtained is the optimal solution of the BACAP model searched by the ETMA algorithm.

[0133] The algorithm ETMA proposed in this invention is compared with the current mainstream berth allocation and quay crane scheduling optimization algorithms. The comparison algorithms include Hybrid Genetic Algorithm (HGA), Genetic Algorithm with the Superiority of Feasible solutions strategy (GA-SF), and Enhanced Non-dominated Sorting Genetic Algorithm II (ENSGA-II).

[0134] The superior performance of the present invention is specifically illustrated by comparing the performance of the present invention with the above three algorithms in examples. The algorithm input parameters are set as follows: population size pop_num = 100, maximum number of iterations iteration_num = 2000, crossover rate crossover_rate = 0.8, mutation rate muation_rate = 0.4, cluster number K = 4, diversity enhancement interval generation G = 50. However, since there is no public standard data set for BACAP at present, the present invention will refer to the existing related technical experimental parameter settings and real port data to provide BACAP simulation data with 40 ships.

[0135] Next, Table II gives the running results of the four algorithms on the instance. The contents of the table are as follows: The first column of the table is the name of the data set V-Data, where V represents the data set size, expressed in the number of ships, Data is the data set number, Best, Mean, and Std are the best value, average value, and standard deviation of the 20 running results of each algorithm, respectively. Here, since ENSGA-II and GA-SF require a large number of iterations to find a better solution, while ETMA and GA converge earlier, in order to be fair, the convergence time of the algorithms is compared. The optimal Best and Mean in each instance are bolded in the table. In the results of the Wilcoxon rank sum test between ETMA and the comparison algorithm, the significance level is taken as 0.05. + / - / ≈ indicates the number of instances where the corresponding ETMA results are significantly better, significantly worse, or have no significant difference than the comparison algorithm under the same data set.

[0136] Table II: Comparison of the results of four algorithms on large-scale datasets

[0137]

[0138] From the experimental results in Table II, we can see that ETMA has obtained the best value that is better than the other three algorithms on all data sets in large-scale problems, and the average solution is also significantly better than GA, GA-SF and ENSGA-II. In addition, in these instances, the average value of ETMA's results is better than the optimal value of the other three algorithms. At the same time, ETMA has a smaller standard deviation and more stable performance. This shows that ETMA has an unignorable advantage in the collaborative optimization of continuous berth allocation and time-varying allocation of quay cranes for the coordinated objectives of ship offset distance and waiting time.

[0139] Figure 6 The figure is a schematic diagram of the convergence speed of the algorithm. The horizontal axis plots the iteration number and the vertical axis plots the fitness value of the optimal individual. It can be seen that compared with the comparative algorithm, the algorithm proposed in the present invention can converge quickly.

[0140] Figure 7 The final solution is visualized, where the horizontal axis represents the shoreline, the vertical axis represents time, and each solid rectangular block represents the berthing situation of a ship (the horizontal axis in the lower left corner of the rectangle represents the berthing position, the vertical axis represents the berthing time, the width of the rectangle represents the length of the ship, and the height represents the working time). The three numbers in the rectangle represent the ship number, the ship offset distance, and the waiting time. The dotted box represents the safe distance of the ship (10% of the length of the ship), and the black dot represents the estimated arrival time and target berth of each ship. The arrow can reflect the size of the ship offset target value. Figure 6It can be seen that the present invention can find an excellent solution for the integrated problem of continuous berth allocation and time-varying quay crane allocation (the fitness value of the solution in this embodiment is 2943). In the figure, most ships can berth at the target berth within the expected waiting time, and those that cannot berth at the target berth within the expected waiting time can also wait for a shorter time or berth near the target berth. The solution given by the algorithm is overall excellent and reasonable, and can effectively solve examples of such problems.

[0141] The present invention also provides a coordinated allocation device for continuous berths and time-varying quay cranes at a wharf, which is used to control the efficiency and highly coordinate the wharf, establish a coordinated optimization problem (BACAP) of continuous berth allocation and time-varying allocation of quay cranes for the joint objective of the ship's offset distance and waiting time, and design an enhanced three-stage memetic algorithm (ETMA) based on the problem model, so that continuous berths can be truly processed. It should be noted that the implementation principle and specific implementation method of the coordinated allocation device for continuous berths and time-varying quay cranes at a wharf can refer to the above-mentioned coordinated allocation method for continuous berths and time-varying quay cranes at a wharf, which will not be repeated below.

[0142] Specifically, the coordinated allocation device for continuous berths and time-varying quay cranes of a wharf includes:

[0143] The ship coding module is used to sort and code each ship according to its serial number to indicate the berthing order of the ship;

[0144] The first stage decoding module is used to obtain a rough feasible initial solution of the first stage according to the berthing sequence of the ships, and the initial solution includes the berth allocated to each ship and the starting berthing time;

[0145] The second stage decoding module is used to use the initial solution obtained by the first stage decoding module as input, allocate a quay crane to each ship in a time-varying manner according to the berth allocation situation, update the ship's starting berthing time and working time, and output the second stage solution;

[0146] The third stage decoding module is used to improve the solution obtained by the second stage decoding module, optimize the berth of the ship, and output the final solution;

[0147] An initial population generation module is used to cluster the ships according to the latest target berths and estimated arrival times of the ships, and then generate an initial population based on the clustering results;

[0148] A population evolution module, used for iterating the evolution of the initial population based on the fitness function;

[0149] The individual optimization module is used to perform local search operations on the best individuals of each generation of population based on the preset local search operator, and to improve and optimize the best individuals of each generation of population;

[0150] The elite retention module is used to apply the elite retention strategy to retain the best preset individuals in each generation to the next generation. When the preset iteration termination condition is reached, the optimal individual that is the optimal solution of the preset model is obtained.

[0151] In summary, the method and device for collaborative allocation of continuous berths and time-varying quay cranes for a wharf provided by the present invention, by executing steps S10-S40, while allocating berths to ships, also achieves time-varying allocation of quay cranes to ships, thereby achieving the purpose of maximizing the utilization rate of quay cranes and improving the efficiency of wharf operations. Then, by executing steps S50-S80, the ships are clustered according to the spatiotemporal data of the estimated arrival time of the ships and the target berths, and diverse and potential individuals are generated according to the clustering results. Then, the population is subjected to crossover and mutation operators designed based on the clustering results, and the target-guided local search operator is executed. Finally, the elite protection strategy is adopted to retain excellent individuals and generate offspring, so that the ships in the problem instance are clustered according to the estimated arrival time and the target berth. The population initialization method and evolution operator based on the clustering results effectively narrow the search range of excellent solutions, avoid meaningless searches, and improve the search efficiency, so that the algorithm can still quickly find a convergent solution when processing large-scale instances. At the same time, the target-guided local search operator is further optimized on the basis of finding an excellent solution to ensure the excellence of the solution.

[0152] Those skilled in the art can clearly understand that, for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example for illustration. In practical applications, the above-mentioned function allocation can be completed by different functional units and modules as needed, that is, the internal structure of the system can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiment can be integrated in a processing unit, or each unit can exist physically separately, or two or more units can be integrated in one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of this application. The specific working process of the units and modules in the above-mentioned system can refer to the corresponding process in the aforementioned method embodiment, which will not be repeated here.

[0153] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described or recorded in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0154] Those of ordinary skill in the art will appreciate that the units and method steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0155] In the embodiments provided by the present invention, it should be understood that the disclosed systems or devices / terminal equipment and methods can be implemented in other ways. For example, the system or device / terminal equipment embodiments described above are only schematic. For example, the division of the modules or units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, and the indirect coupling or communication connection of the system or unit can be electrical, mechanical or other forms.

[0156] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0157] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0158] The present invention is not limited to what is described in the specification and implementation modes, and therefore additional advantages and modifications can be easily realized by those skilled in the art. Therefore, without departing from the spirit and scope of the general concept defined by the claims and their equivalents, the present invention is not limited to the specific details, representative devices, and illustrative examples shown and described herein.

Claims

1. A method for coordinating the allocation of continuous berths and time-varying quay cranes at a wharf, characterized in that Considering the waiting time of the ship and the offset distance of the target berth, the ship is allocated berths on the shoreline of continuous berths and the quay crane is allocated in a time-varying manner, and such large-scale problems can be solved. The specific steps include: Step S10: sorting and coding each ship according to the ship serial number to indicate the berthing order of the ships; Step S20: obtaining a rough feasible initial solution of the first stage according to the berthing sequence of the ships, the initial solution including the berth allocated to each ship, the starting berthing time and the working time; Step S30: using the initial solution obtained in step S20 as input, allocating a quay crane to each ship in a time-varying manner according to the berth allocation situation, updating the ship's starting berthing time and working time, and outputting a second-stage solution; Step S40: further improving the solution obtained in step S30, optimizing the berth of the ship, and outputting a final solution; Step S50: clustering the ships according to their target berths and estimated arrival times, and then generating an initial population based on the clustering results; Step S60: performing evolution iteration on the initial population based on a preset fitness function; Step S70: performing a local search operation on the best individual of each generation of population using a preset local search operator to improve and optimize the best individual of each generation of population; Step S80: Applying the elite retention strategy, retaining the best preset individuals in each generation to the next generation, and obtaining the optimal individual that is the optimal solution of the preset model when the preset iteration termination condition is reached; The step S20 comprises: The working hours of the corresponding ships are calculated according to the allocated preset number of quay cranes, and then the berths and starting berthing time are allocated to each ship in turn according to the berthing order of the ships; Specifically, it is determined whether the ship i overlaps with the moored ships in time and space. If so, feasible berthing segments are searched in chronological order; The step S30 comprises: Step S31: Divide the time periods according to the initial solution obtained in step S20 so that the number of ships at the dock does not change in each time period, and obtain the starting timetable time_period of each time period; Step S32: at the beginning of each time period, a quay crane is allocated to each ship according to the proportion of the remaining workload of the ships at the dock, and the number of quay cranes must also meet the limit; Step S33: If the ship i can complete the work within the next time period, the berthing time of the ship i is shortened and the departure time of the ship i is updated; Step S34: searching for a ship h whose berth overlaps with that of the ship i, and accordingly advancing the berthing time of the ship h to shorten the waiting berthing time of the ship h, and updating the start schedule time_period; The step S40 comprises: Detect the berthing position of each ship and move it horizontally to get it as close to its target berth as possible.

2. The method for coordinating the continuous berths and time-varying quay cranes for a wharf according to claim 1, characterized in that: The step S50 comprises: Step S51: using a preset clustering operator to divide the ship set into K categories, and sorting the ships according to their estimated arrival times; Step S52: on the basis of the ordering by estimated arrival time, the order of the ships in the same cluster after clustering is randomly sorted to obtain different individuals; Step S53: Generate an initial population based on the result of step S52.

3. The method for coordinating the continuous berths and time-varying quay cranes for a wharf according to claim 1, characterized in that: The step S60 comprises: Step S61: selecting a parent generation using a roulette algorithm according to a value of a preset fitness function; Step S62: randomly selecting a cluster from the clustering results by using a preset crossover operator, then randomly selecting two crossover points in the cluster, and sequentially crossovering the two selected parent generations to be crossovered; Step S63: randomly selecting two clusters from the clustering results through a preset mutation operator, and randomly selecting two points in the two clusters for exchange to complete the mutation of the individual.

4. The method for coordinating the continuous berths and time-varying quay cranes for a wharf according to claim 1, characterized in that: In step S70, the following operations are performed on the ship in descending order of the objective function value: Step S71: Check whether the ship can directly berth at the target berth at the estimated arrival time. If there is an overlap with other ships, search for a feasible berth section in the current time period; Step S72: while keeping the berthing time unchanged, check whether the idle quay cranes in the current period are sufficient for the ship to complete the work; if not, try to extend the berthing time of the ship, otherwise go to step S73; Step S73: traverse from the estimated arrival time to the allocated parking time, and check whether there is a parking time for the berth segment that can be parked. If so, repeat the above steps S71-S72, if not, stop.

5. The method for coordinating the continuous berths and time-varying quay cranes for a wharf according to claim 1, characterized in that: Also includes the steps: Determine whether the number of iterations is a multiple of G. If the result is yes, re-cluster; if the result is no, retain the elite individuals in the population; where G is a preset positive integer and G≥2.

6. The method for coordinating the continuous berths and time-varying quay cranes for a wharf according to claim 1, characterized in that: In step S60, the fitness function is set to the inverse of the objective function; wherein the objective function is: V is the set of ships, V={1,2,…,|V|}, a i is the estimated arrival time of ship i, ts i is the start berthing time of ship i, b i is the target berth position of ship i, p i is the berthing position of ship i, α, β are the weight values ​​of the preset model target.

7. A coordinated allocation device for continuous berths and time-varying quay cranes at a wharf, characterized in that: include: The ship coding module is used to sort and code each ship according to its serial number to indicate the berthing order of the ship; The first stage decoding module is used to obtain a rough feasible initial solution of the first stage according to the berthing sequence of the ships, and the initial solution includes the berth allocated to each ship and the starting berthing time; The second stage decoding module is used to use the initial solution obtained by the first stage decoding module as input, allocate a quay crane to each ship in a time-varying manner according to the berth allocation situation, update the ship's starting berthing time and working time, and output the second stage solution; The third stage decoding module is used to improve the solution obtained by the second stage decoding module, optimize the berth of the ship, and output the final solution; An initial population generation module is used to cluster the ships according to their target berths and estimated arrival times, and then generate an initial population based on the clustering results; The population evolution module is used to iterate the evolution of the initial population based on the fitness function; The individual optimization module is used to perform local search operations on the best individuals of each generation of population based on a preset local search operator, and to improve and optimize the best individuals of each generation of population; The elite retention module is used to apply the elite retention strategy to retain the best preset individuals in each generation to the next generation. When the preset iteration termination condition is reached, the optimal individual that is the optimal solution of the preset model is obtained; The first stage decoding module is also used for: The working hours of the corresponding ships are calculated according to the assigned preset number of quay cranes, and then the berths and the start berthing time are allocated to each ship in turn according to the berthing order of the ships; specifically, it is determined whether the ship i overlaps with the berthed ships in time and space. If there is an overlap, the feasible berthing segments are searched in time order; The second stage decoding module is also used for: Divide the time period according to the initial solution obtained by the decoding module in the first stage so that the ships at the dock do not change in each time period, and obtain the starting timetable time_period of each time period; at the beginning of each time period, allocate a quay crane to each ship according to the proportion of the remaining workload of the ships at the dock, and the quay crane quantity range limit must be met at the same time; if the ship i can complete the work in the next time period, shorten the berthing time of the ship i, and update the departure time of the ship i; search for the ship h that has a berth overlapping with the ship i, and advance the berthing time of the ship h accordingly, so as to shorten the waiting berthing time of the ship h, and update the starting timetable time_period; The third stage decoding module is also used for: Detect the berthing position of each ship and move it horizontally to get it as close to its target berth as possible.

Citation Information

Patent Citations

  • Automatic container wharf berth and quay crane optimal distribution method

    CN115713170A

  • Wharf berth and quay crane distribution optimization method based on genetic algorithm

    CN116611657A