Container loading and unloading sequence decision-making prospective and adaptive optimization method and system

Abstract

The invention discloses a prospective and adaptive optimization method and system for container loading and unloading sequence decision, and the method comprises the steps: calculating the total profit and loss time of candidate operation based on the operation state of a container; performing local optimal selection based on the total profit and loss time; based on the local optimal selection, executing the selected operation; and performing dynamic optimization of the sequence according to deviation or interruption in real-time operation. The method is high in adaptability, any operation interruption or deviation can be evaluated and adjusted in real time, and the method can be used for intelligent decision making of a complete sequence in an operation planning stage by repeatedly calling a heuristic algorithm process and can also be suitable for dynamic decision making based on an actual operation condition in a real-time scheduling control stage; and the global optimization is good, and the future trajectory profit-loss time and the future landing profit-loss time are taken into evaluation, so that the step-by-step decision process has a prospective view, and the sequence decision optimization effect is systematically improved.

Description

Technical Field

[0001] The present invention relates to the field of scheduling of port machinery for container stuffing and discharging, and particularly to a method and system for prospectively and adaptively optimizing the decision-making of container loading and unloading sequences. Background Art

[0002] In the operation of port machinery for container loading and unloading in container ports, due to the strict requirements for safety and efficiency, the container loading and unloading sequence is generally pre-determined by operation planners in advance and manually controlled by on-site operation commanders during real-time operations. However, the realization of intelligent decision-making and automatic control of port machinery operations is an inevitable trend in the future development of container ports. This process mainly includes the following three specific requirements: ensuring the safety and convenience of spreader operations, that is, considering the influence of the container bay shape on the spreader movement trajectory and operation difficulty; improving the loading and unloading operation efficiency, that is, the port machinery spreader completes the container sequence loading and unloading operation in a shorter time, including the support for the operation process of loading and unloading on the same bay side; having the ability to dynamically optimize the sequence order, that is, adaptively adjusting the loading and unloading sequence plan according to the deviation in the real-time operation process.

[0003] The existing technologies for container loading and unloading sequence decision-making mainly rely on fixed rules and manual adjustment methods, that is, the loading and unloading operation plan is made according to the direction from the sea side to the land side or from the land side to the sea side, and then the sequence is fine-tuned based on the real-time operation process. This method can be well applied to the optimization of the separation of loading and unloading sequences, but it is not applicable to the situation where the port machinery loads and unloads on the same bay side, that is, unloading a container from a ship while loading another container, because the optimal operation direction may change at any time due to the alternation of loading and unloading. Therefore, the existing container loading and unloading sequence decision-making methods have three major drawbacks: relying on manual decision-making control and lacking foresight, that is, the degree of intelligence is not high; being unable to support operation processes that can improve loading and unloading efficiency such as loading and unloading on the same bay side; being unable to perform efficient adaptive optimization based on fixed rules. Existing research has not yet had a sequence decision optimization method for the shortest loading and unloading operation time based on the spreader trajectory of quay cranes (considering operation safety and convenience), resulting in great limitations in the application of the current loading and unloading sequence decision-making algorithms. Summary of the Invention

[0004] Aiming at the lack of consideration of the spreader trajectory and landing environment of container port machinery in the existing loading and unloading sequence decision-making, as well as the three major drawbacks that still exist, namely, low intelligence, non-support for loading and unloading on the same bay side, and inability to perform adaptive optimization, the present invention provides a practical and prospective adaptive sequence optimization method system for the sequence decision-making of ship loading and unloading operations in container terminals. The two major technical problems mainly solved by this invention are: how to evaluate the sequence decision based on the control of the quay crane spreader trajectory, and how to ensure that the sequence decision based on the local area has global optimization.

[0005] To achieve the above object, the present invention provides a prospective and adaptive optimization method for the container handling sequence decision-making, and the steps include:

[0006] Based on the operation status of the container, calculate the total profit and loss time of the candidate operation;

[0007] Based on the total profit and loss time, make a local optimal selection;

[0008] Based on the local optimal selection, execute the selected operation;

[0009] According to the deviation or interruption in the real-time operation, perform dynamic optimization of the sequence.

[0010] Preferably, the expression of the total profit and loss time is as follows:

[0011]

[0012] Among them, B s represents the total profit and loss time caused by taking the loading / unloading operation of the container at position s as the next operation; represents the current trajectory profit and loss time brought to the empty movement by taking the loading / unloading operation of the container at position s as the next operation; represents the current trajectory profit and loss time brought to the heavy movement by taking the loading / unloading operation of the container at position s as the next operation; represents the current landing profit and loss time caused by taking the loading / unloading operation of the container at position s as the next operation; represents the future trajectory profit and loss time that will inevitably be brought to the subsequent operation process after completing the loading / unloading operation of the container at position s; represents the future landing profit and loss time that will inevitably be brought to the subsequent operation process after completing the loading / unloading operation of the container at position s.

[0013] Preferably, the calculation methods of the and include:

[0014]

[0015] Among them, r and s represent the positions of the containers in the ship's hold, and W D represents the set of the positions of the containers unloaded from the ship; W L represents the set of the positions of the containers loaded onto the ship; X rs =1 indicates that the container at s is operated immediately after the container at r; i0 is the lowest ship's hold section, that is, the ship's hold section when all the loaded and unloaded containers are emptied; i s is the ship's hold section when loading / unloading the container at position s; represents at the current ship's hold section i sThe time of the empty spreader movement trajectory from the truck position o to the container position s below; Indicates the waiting time for releasing the previous discharged container from the truck position and grasping the next loaded container; Indicates at the current ship bay section i s The time of the empty spreader movement trajectory from the container position r to the container position s below; Indicates at the current ship bay section i s The time of the empty spreader movement trajectory from the container position r to the truck position o below; Indicates at the current ship bay section i s The time of the loaded spreader movement trajectory from the container position s to the truck position o below; Indicates the time of the loaded spreader movement trajectory from the container position s to the truck position o at the lowest ship bay section i0; Indicates at the current ship bay section i s The time of the loaded spreader movement trajectory from the truck position o to the container position s below; Indicates the time of the quay crane's loaded spreader movement trajectory from the truck position o to the container position s at the lowest ship bay section i0; Indicates at the current ship bay section i s The landing time of the empty spreader when grasping the container located at s below; Indicates at the current ship bay section i s The landing time of the loaded spreader when placing the container located at s below; t L1 Indicates the shortest landing time when the loaded container needs to be placed and landed in the ship bay and there is an adjacent container to rely on; t D1 Indicates the shortest landing time when the discharged container needs to be grasped and landed in the ship bay and the landing guide plate can be lowered or there is an adjacent container to rely on.

[0016] Preferably, the and calculation methods include:

[0017]

[0018] where γ represents the position of the container in the ship bay; Y sγ =1 indicates that the container at γ is operated after the container at s; i s is the ship bay section before loading / unloading the container at position s, and i s’ is the ship bay section after loading / unloading the container at position s; Indicates the time of the loaded spreader movement trajectory from the truck position o to the container position r at the ship bay section after the container at s is loaded / unloaded; It represents the time of the heavy load movement trajectory of the spreader from the position o of the straddle carrier under the ship's shell section before the container located at s is loaded and unloaded to the position r of the container. It represents the time when the heavy load spreader lands when placing the container located at r under the ship's shell section after the container located at s is loaded and unloaded. It represents the time when the heavy load spreader lands when placing the container located at r under the ship's shell section before the container located at s is loaded and unloaded. It represents the time when the empty load spreader lands when grasping the container located at r under the ship's shell section after the container located at s is loaded and unloaded. It represents the time when the empty load spreader lands when grasping the container located at r under the ship's shell section before the container located at s is loaded and unloaded.

[0019] Preferably, the implementation of the local optimal selection includes: a heuristic-based greedy algorithm, which determines the loading or unloading task of the next job through local decisions at each step, and finally forms a complete container loading and unloading sequence plan.

[0020] The present invention also provides a forward-looking and adaptive optimization system for container loading and unloading sequence decision-making. The system is used to implement the above method and includes: a calculation module, a selection module, an execution module, and an optimization module;

[0021] The calculation module is used to calculate the total profit and loss time of candidate jobs based on the operation status of the container.

[0022] The selection module is used to make a local optimal selection based on the total profit and loss time.

[0023] The execution module is used to execute the selected job based on the local optimal selection.

[0024] The optimization module is used to perform dynamic optimization of the sequence according to the deviation or interruption in real-time operations.

[0025] Preferably, the expression of the total profit and loss time is as follows:

[0026]

[0027] Where B s represents the total profit and loss time caused by taking the loading / unloading operation of the container at position s as the next job; represents the current trajectory profit and loss time brought to the empty load movement by taking the loading / unloading operation of the container at position s as the next job; represents the current trajectory profit and loss time brought to the heavy load movement by taking the loading / unloading operation of the container at position s as the next job; represents the current landing profit and loss time caused by taking the loading / unloading operation of the container at position s as the next job; It represents the future trajectory profit and loss time that will inevitably be brought to the subsequent operation process after completing the loading / unloading operation at the container position s. It represents the future landing profit and loss time that will inevitably be brought to the subsequent operation process after completing the loading / unloading operation at the container position s.

[0028] Preferably, the working process of the selection module includes: based on the heuristic greedy algorithm, determining the loading or unloading task of the next operation through local decisions at each step, and finally forming a complete container loading and unloading sequence plan.

[0029] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0030] The present invention has high adaptability, and any operation interruption or deviation can be evaluated and adjusted in real time. It can be used for intelligent decision-making of the entire sequence in the operation planning stage by repeatedly calling the heuristic algorithm process, and is also applicable to dynamic decision-making based on the actual operation situation in the real-time scheduling control stage; it has good global optimization. Since the future trajectory profit and loss time and the future landing profit and loss time are incorporated into the evaluation, the step-by-step decision-making process has a forward-looking perspective, thereby systematically improving the optimization effect of sequence decision-making. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] In order to more clearly illustrate the technical solutions of the present invention, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0032] Figure 1 It is a combination of the heavy-load movement trajectory and the empty-load movement trajectory under each operation type in the embodiment of the present invention; among them, (a) shows that there is only a loading operation within one movement cycle of the spreader, (b) shows that there is only an unloading operation within one movement cycle of the spreader, and (c) shows that there are both loading and unloading operations within one movement cycle of the spreader;

[0033] Figure 2 It is a schematic diagram of each landing scenario under the control of the spreader in the embodiment of the present invention;

[0034] Figure 3 It is a schematic diagram of trajectory profit and loss and landing profit and loss in the embodiment of the present invention. Among them, (a) shows the trajectory profit and loss caused by loading the container; (b) shows the landing profit and loss caused by loading the container;

[0035] Figure 4 It is a schematic diagram of adaptive optimization in the method in the embodiment of the present invention;

[0036] Figure 5Schematic diagram of operation sequence decision-making during the separation of loading and unloading in the embodiments of the present invention; wherein (a) represents the unloading sequence decision; (b) represents the loading sequence decision;

[0037] Figure 6 Schematic diagram of operation sequence decision-making for simultaneous side loading and unloading in the embodiments of the present invention. Detailed implementation manners

[0038] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0039] To make the above objects, features, and advantages of the present invention more obvious and understandable, the present invention will be further described in detail below in conjunction with the accompanying drawings and specific implementation manners.

[0040] Embodiment 1

[0041] This embodiment provides a prospective and adaptive optimization method for container loading and unloading sequence decision-making, and the steps include:

[0042] S1. Calculate the total profit and loss time of candidate operations based on the operation status of the container.

[0043] First, discuss how to evaluate each candidate operation, and this evaluation will be based on the profit and loss time. The profit and loss time can reflect the increase or decrease in operation time brought by the current candidate operation during execution, that is, the current profit and loss time, and the increase or decrease in operation time that the current candidate operation will inevitably bring to future operations after execution, that is, the future profit and loss time. Among them, the former compares the advantages and disadvantages of the current operation candidate in the current operation state, while the latter compares the impact of the current operation candidate on the future operation state. For the current profit and loss time and the future profit and loss time, they are both composed of two parts: trajectory profit and loss and positioning profit and loss. The former evaluates the impact of the ship-beam cross-section on the trajectory time based on the curve trajectory of the quay crane spreader, and the latter evaluates the impact of the positioning environment on the positioning time based on the swing suppression and blocking of the quay crane spreader. Since different loading and unloading ship sequences will result in different trajectory profit and loss and positioning profit and loss being quantitatively calculated to support the optimization of sequence decision-making. Finally, the profit and loss time can be divided into five main parts: the current trajectory profit and loss time of no-load movement, the current trajectory profit and loss time of heavy-load movement, the current positioning profit and loss time, the future trajectory profit and loss time, and the future positioning profit and loss time. Table 1 shows the definition of the parameters required for calculating the profit and loss time.

[0044] Table 1

[0045]

[0046] The calculation of the profit and loss time needs to be based on a certain operation time standard, that is, only the increased or decreased part relative to a good operation time is calculated. However, the time standards for the above five kinds of profit and loss time are not the same. The specific calculation process is as follows:

[0047] The comparison standard for the current trajectory profit and loss time of the heavy-load movement is the shortest heavy-load trajectory time, that is, the heavy-load trajectory time when the ship's shell profile is the lowest. The current trajectory profit and loss time of the heavy-load movement is the difference between the current trajectory time of the heavy-load movement (based on the ship's shell profile i s ) and the shortest heavy-load trajectory time. As Figure 1 shown, for the discharged containers, the heavy-load movement is from the ship's shell position s to the truck position o; while for the loaded containers, the heavy-load movement is from the truck position o to the ship's shell position s. The calculation formula is as follows:

[0048]

[0049] Among them, s represents the position of the container in the ship's shell, W D represents the set of positions of the discharged containers; W L represents the set of positions of the loaded containers; i0 is the lowest ship's shell profile, that is, the ship's shell profile when all the loaded and discharged containers are emptied; i s is the ship's shell profile when loading / unloading the container at position s; represents the spreader heavy-load movement trajectory time from the container position s to the truck position o under the current ship's shell profile i s ; represents the spreader heavy-load movement trajectory time from the container position s to the truck position o under the lowest ship's shell profile i0; represents the spreader heavy-load movement trajectory time from the truck position o to the container position s under the current ship's shell profile i s ; represents the quay crane spreader heavy-load movement trajectory time from the truck position o to the container position s under the lowest ship's shell profile i0.

[0050] For the current trajectory profit and loss time of the no-load movement, due to the allowed operation process of loading and discharging simultaneously on the same side of the ship's shell, the optimal no-load trajectory time can be the no-load trajectory time between two relatively close container positions in the ship's shell, that is, from the ship's shell position of the previous loaded container to the ship's shell position of the next discharged container. Its shortest time can be close to zero. Therefore, the actual no-load movement trajectory time is directly used as the current trajectory profit and loss time of this no-load movement. As Figure 1As shown in the figure, for the discharging containers whose previous operation was discharging, the empty movement is from the container yard position o to the ship's hold position s; for the discharging containers whose previous operation was loading, the empty movement is from the previous loading ship's hold position r to the current discharging ship's hold position s; for the loading containers whose previous operation was discharging, the empty movement is the waiting time at the container yard position o for the next loading container; for the loading containers whose previous operation was loading, the empty movement is from the previous loading ship's hold position r to the container yard position o. The calculation formula is as follows:

[0051]

[0052] Among them, X rs = 1 means that the container at s is operated immediately after the container at r; represents the empty movement trajectory time of the spreader from the container yard position o to the container position s under the current ship's hold section i s ; represents the waiting time for releasing the previous discharging container and grasping the next loading container at the container yard position; represents the empty movement trajectory time of the spreader from the container position r to the container position s under the current ship's hold section i s ; represents the empty movement trajectory time of the spreader from the container position r to the container yard position o under the current ship's hold section i s .

[0053] For the current landing profit and loss time, the landing time on the container yard is not affected by the loading and unloading sequence and is not taken into consideration, and only the landing time of the containers inside the ship's hold is considered. As Figure 3 shown, there are some optimal landing environments that minimize the landing time. Among them, the loading containers need to be placed and landed inside the ship's hold. When there are adjacent containers to rely on, as Figure 2 (L1) shown, the spreader shaking can be quickly stabilized, minimizing the landing time, denoted as t L1 ; the discharging containers need to be grabbed and landed inside the ship's hold. When the landing guide plate can be lowered or there are adjacent containers to rely on, as Figure 2 (D1) shown, the spreader shaking can be quickly stabilized, minimizing the landing time, denoted as t D1 . Then when the actual landing time is higher than this shortest time, it will result in an increase in the potential operation time. The calculation formula is as follows:

[0054]

[0055] Among them, represents the landing time of the empty spreader when grasping the container at s under the current ship's hold section i s ; represents the landing time of the empty spreader when grasping the container at s under the current ship's hold section is The time it takes for the heavy-load spreader to land when placing the container at s.

[0056] The future profit and loss time is the key manifestation of the forward-looking nature of this embodiment. By quantitatively evaluating the impact of the current choice on the future operation time, the key evaluation factors of global optimization can be incorporated on the basis of local optimization under the greedy algorithm, so that the current decision promotes the improvement of the operation efficiency of the entire operation process. Before and after the container at position s is operated, there are two types of ship shell profiles, namely i s and i s’ In this embodiment, the operation time caused by the two ship shell profiles is compared for all containers that will be operated later, and the operation time is calculated based on i s’ The future operation time caused by this is the standard: i s The part that is bound to increase or decrease is treated as future profit or loss time.

[0057] For future trajectory gain and loss time, such as Figure 3 As shown in (a), after the red box is loaded, the heavy load motion trajectory of the box on the sea side of the box will inevitably be raised, which will inevitably increase the total operation time. On the contrary, the operation trajectory of the box on the land side of the box will not be affected. For the future landing gain time, Figure 3 As shown in (b), when the red box is loaded, the presence of the blocking box will inevitably lead to an inevitable increase in the placement time. In this way, the inevitable impact of the current decision on the future operation time will be quantified by the following formula:

[0058]

[0059] Where γ represents the position of the container in the ship; Y sγ =1 means that the container at γ is operated after the container at s; i s is the section of the ship before loading / unloading the container at position s, and i s’ The section of the ship shell after loading / unloading the container at position s; It represents the heavy-load movement trajectory time of the spreader from the truck position o to the container position r under the ship shell section after the container at s is loaded and unloaded; represents the heavy-load movement trajectory time of the spreader from the truck position o to the container position r under the ship shell section before the container at s is loaded or unloaded; It represents the time for the heavy load spreader to be placed when placing the container at r under the ship shell section after the container at s is loaded and unloaded; It represents the time for the heavy load spreader to be placed when placing the container at r under the ship shell section before the container at s is loaded or unloaded; It represents the landing time of the empty spreader when grabbing the container at r under the ship shell section after the container at s is loaded and unloaded; It represents the time it takes for the empty spreader to land when grabbing the container at r under the ship shell section before the container at s is loaded or unloaded. Then, the total profit and loss time is the sum of the above five profit and loss times, as shown in the following formula:

[0060]

[0061] S2. Based on the total profit and loss time, perform local optimal selection.

[0062] This container loading and unloading sequence decision method is based on a heuristic greedy algorithm (GreedyAlgorithm), which determines the loading or unloading task of the next operation through local decisions at each step, and finally forms a complete container loading and unloading sequence plan. Different from the traditional greedy algorithm that only evaluates candidates based on the current state, this method system, based on the current state evaluation, also prospectively considers the impact of the current decision on the future state and decision (Look-ahead), so it has the characteristics of global decision optimization. The process and formula of the heuristic greedy algorithm of this embodiment are shown in Table 2.

[0063] Table 2

[0064]

[0065] S3. Based on the local optimal selection, execute the selected task.

[0066] S4. Dynamically optimize the sequence based on deviations or interruptions in real-time operations.

[0067] Based on the above evaluation of profit and loss time and the greedy algorithm process, the forward-looking container loading and unloading sequence decision adaptive optimization algorithm process can be referred to Figure 4 .

[0068] Embodiment 2

[0069] The following is combined with this implementation to verify the advancement of the present invention compared to the prior art.

[0070] 1. When only unloading or loading containers is done, take six containers to be unloaded as an example. Figure 5(a). Initially, only containers a, b, and c can be unloaded from the ship. By calculating their total profit and loss times according to equations (1)-(6) respectively, it is found that the total profit and loss time of container a is the smallest. That is, container a is listed as the first container to be unloaded from the ship. After updating the ship's profile, containers d, b, and c can be unloaded. Then, calculate their total profit and loss times, and container d is taken as the second container to be unloaded. And so on, the loading and unloading sequences of all containers can be determined. When there is only loading operation of containers, the decision-making process can refer to Figure 5 (b), which is similar to the decision-making process of the unloading sequence.

[0071] 2. When loading and unloading simultaneously on the same side of the ship's side, the decision-making process is as shown in Figure 6 the example. The profit and loss time is calculated according to the formulas given in equations (1)-(6). When the container for the third operation has been determined, the hatch cover 1 (the leftmost one) will be opened. Then, the containers that can be the next operation include the uppermost container in the hold under hatch cover 1 and the uppermost containers on the upper deck of hatch covers 2 and 3 (the right side). By comparing the total profit and loss time values, the container for the fourth operation can be determined as a loading container shown in the figure. Continue to update the ship's profile and the containers that can be operated currently. By comparing the total profit and loss time values, the container for the fifth operation is determined as an unloading container shown in the figure. And so on, the loading and unloading order of the subsequent containers can be determined. This sequence adopts the loading and unloading process of loading and unloading alternately on the same side of the ship's side.

[0072] Table 3

[0073]

[0074] The comparison test with manual decision-making in production practice is shown in Table 3. It can be found that under this data set, compared with manual decision-making, the decision-making method of the present invention increases the average number of loading and unloading simultaneously on the same side by 12 times, reduces the total operation time by 3.09%, among which the no-load movement time is reduced by 10.60%, the heavy-load movement time increases slightly by 0.95%, and the positioning time is reduced slightly by 0.54%. The average calculation response time of the decision-making method of the present invention is 1.705 s.

[0075] Embodiment 3

[0076] The present invention also provides a forward-looking and adaptive optimization system for container loading and unloading sequence decision-making, including: a calculation module, a selection module, an execution module, and an optimization module; the calculation module is used to calculate the total profit and loss time of candidate operations based on the operation status of containers; the selection module is used to make a locally optimal selection based on the total profit and loss time; the execution module is used to execute the selected operation based on the locally optimal selection; the optimization module is used to perform dynamic optimization of the sequence according to the deviation or interruption in real-time operations.

[0077] Next, in combination with this embodiment, it will be described in detail how the present invention solves the technical problems in actual work.

[0078] First, the calculation module calculates the total profit and loss time of the candidate operation based on the operation status of the container.

[0079] First, discuss how to evaluate each candidate operation, and this evaluation will be based on the profit and loss time. The profit and loss time can reflect the increase or decrease in operation time brought by the current candidate operation during execution, that is, the current profit and loss time, and the increase or decrease in operation time that the current candidate operation will inevitably bring to future operations after execution, that is, the future profit and loss time. Among them, the former compares the advantages and disadvantages of the current operation candidate under the current operation status, while the latter compares the impact of the current operation candidate on the future operation status. For the current profit and loss time and the future profit and loss time, they are both composed of two parts: trajectory profit and loss and landing profit and loss. The former evaluates the impact of the ship-beam section on the trajectory time based on the curve trajectory of the quay crane spreader, and the latter evaluates the impact of the landing environment on the landing time based on the swing suppression and blocking of the quay crane spreader. Due to different loading and unloading ship sequences, different trajectory profit and loss and landing profit and loss will be quantitatively calculated to support the optimization of sequence decision-making. Finally, the profit and loss time can be divided into five main parts: the current trajectory profit and loss time of the empty movement, the current trajectory profit and loss time of the heavy movement, the current landing profit and loss time, the future trajectory profit and loss time, and the future landing profit and loss time. Table 1 shows the definition of the parameters required for calculating the profit and loss time.

[0080] The calculation of the profit and loss time needs to be based on a certain operation time standard, that is, only calculate the increased or decreased part relative to a good operation time, and the time standards on which the above five kinds of profit and loss times are based are not the same. The specific calculation process is as follows:

[0081] The comparison standard for the current trajectory profit and loss time of the heavy movement is the shortest heavy trajectory time, that is, the heavy trajectory time when the ship-beam section is the lowest The current trajectory profit and loss time of the heavy movement is the difference between the current trajectory time of the heavy movement (based on the ship-beam section i s ) and the shortest heavy trajectory time. As Figure 1 shown, for the discharged containers, the heavy movement is from the ship-beam compartment s to the truck position o; while for the loaded containers, the heavy movement is from the truck position o to the ship-beam compartment s. The calculation formula is as follows:

[0082]

[0083] Among them, s represents the position of the container in the ship-beam, W D represents the set of positions of the discharged containers; W L represents the set of positions of the loaded containers; i0 is the lowest ship-beam section, that is, the ship-beam section when all the loaded and unloaded containers are emptied; is The ship's hull profile when loading / unloading the container at position s; Indicates the current ship's hull profile i s The time of the hoist's heavy-load movement trajectory from the container position s to the truck position o under the current ship's hull profile i; Indicates the time of the hoist's heavy-load movement trajectory from the container position s to the truck position o under the lowest ship's hull profile i0; Indicates the current ship's hull profile i s The time of the hoist's heavy-load movement trajectory from the truck position o to the container position s under the current ship's hull profile i; Indicates the time of the quay crane's heavy-load movement trajectory from the truck position o to the container position s under the lowest ship's hull profile i0.

[0084] For the current trajectory profit and loss time of the no-load movement, due to the allowance of the simultaneous loading and unloading operation process on the same hull side, the optimal no-load trajectory time can be the no-load trajectory time between two relatively close container positions within the hull, that is, from the hull berth of the previous loaded container to the hull berth of the next unloaded container. Its shortest time can be close to zero, so the actual no-load movement trajectory time is directly used as the current trajectory profit and loss time of this no-load movement. As Figure 1 shown, for the unloaded container with the previous operation being unloading, the no-load movement is from the truck position o to the hull berth s; for the unloaded container with the previous operation being loading, the no-load movement is from the hull berth r of the previous loaded container to the current hull berth s of the unloaded container; for the loaded container with the previous operation being unloading, the no-load movement is the waiting time at the truck position o for the next loaded container; for the loaded container with the previous operation being loading, the no-load movement is from the hull berth r of the previous loaded container to the truck position o. The calculation formulas are as follows:

[0085]

[0086] where, X rs = 1 indicates that the container at s is operated immediately after the container at r; Indicates the time of the hoist's no-load movement trajectory from the truck position o to the container position s under the current ship's hull profile i s ; Indicates the waiting time for releasing the previous unloaded container and grasping the next loaded container at the truck position; Indicates the current ship's hull profile i s The time of the hoist's no-load movement trajectory from the container position r to the container position s under the current ship's hull profile i; Indicates the current ship's hull profile i s The time of the hoist's no-load movement trajectory from the container position r to the truck position o under the current ship's hull profile i.

[0087] For the current landing time, the landing time on the container truck is not affected by the loading and unloading sequence and is not taken into consideration. Only the container landing time in the ship is considered. Figure 3 As shown in Figure 1, there are some optimal placement environments that minimize the placement time. Among them, when the loading container needs to be placed in the ship, when there are adjacent containers to rely on, such as Figure 2 As shown in (L1), the sling shaking can be quickly stabilized to minimize the landing time, which is recorded as t L1 ; The unloading container needs to be grabbed and placed in the ship's shell. When the positioning guide can be lowered or there is an adjacent container to rely on, Figure 2 As shown in (D1), the sling shaking can be quickly stabilized to minimize the landing time, which is recorded as t D1 . If the actual landing time is higher than the minimum time, the potential operation time will increase. The calculation formula is as follows:

[0088]

[0089] in, Indicates that at the current ship shell section i s The time for the empty spreader to land when grabbing the container at s; Indicates that at the current ship shell section i s The time it takes for the heavy-load spreader to land when placing the container at s.

[0090] The future profit and loss time is the key manifestation of the forward-looking nature of this embodiment. By quantitatively evaluating the impact of the current choice on the future operation time, the key evaluation factors of global optimization can be incorporated on the basis of local optimization under the greedy algorithm, so that the current decision promotes the improvement of the operation efficiency of the entire operation process. Before and after the container at position s is operated, there are two types of ship shell profiles, namely i s and i s’ In this embodiment, the operation time caused by the two ship shell profiles is compared for all containers that will be operated later, and the operation time is calculated based on i s’ The future operation time caused by this is the standard: i s The part that is bound to increase or decrease is treated as future profit or loss time.

[0091] For future trajectory gain and loss time, such as Figure 3 As shown in (a), after the red box is loaded, the heavy load motion trajectory of the box on the sea side of the box will inevitably be raised, which will inevitably increase the total operation time. On the contrary, the operation trajectory of the box on the land side of the box will not be affected. For the future landing gain time, Figure 3 As shown in (b), when the red box is loaded, the presence of the blocking box will inevitably lead to an inevitable increase in the placement time. In this way, the inevitable impact of the current decision on the future operation time will be quantified by the following formula:

[0092]

[0093] Among them, γ represents the position of the container in the ship's bay; Y sγ = 1 indicates that the container located at γ is operated after the container located at s; i s is the ship's bay section before loading / unloading the container at position s, and i s’ is the ship's bay section after loading / unloading the container at position s; represents the time of the heavy-load movement trajectory of the spreader from the undercarriage position o to the container position r under the ship's bay section after the container located at s is loaded / unloaded; represents the time of the heavy-load movement trajectory of the spreader from the undercarriage position o to the container position r under the ship's bay section before the container located at s is loaded / unloaded; represents the landing time of the heavy-load spreader when placing the container at r under the ship's bay section after the container located at s is loaded / unloaded; represents the landing time of the heavy-load spreader when placing the container at r under the ship's bay section before the container located at s is loaded / unloaded; represents the landing time of the empty-load spreader when gripping the container at r under the ship's bay section after the container located at s is loaded / unloaded; represents the landing time of the empty-load spreader when gripping the container at r under the ship's bay section before the container located at s is loaded / unloaded.

[0094] Then, the total profit and loss time is the sum of the above five profit and loss times, as shown in the following formula:

[0095]

[0096] After that, the selection module makes a local optimal selection based on the total profit and loss time.

[0097] This container loading and unloading sequence decision method is based on a heuristic greedy algorithm (Greedy Algorithm), and determines the next loading or unloading task of the ship by making local decisions at each step, finally forming a complete container loading and unloading sequence plan. Different from the traditional greedy algorithm that only evaluates candidates based on the current state, this method system also prospectively considers the impact of the current decision on future states and decisions (Look-ahead) on the basis of the current state evaluation, so it has the characteristics of global decision optimization.

[0098] The execution module executes the selected operation based on the local optimal selection.

[0099] Finally, the optimization module performs dynamic optimization of the sequence according to the deviations or interruptions in the real-time operation.

[0100] Based on the above evaluation of the profit and loss time and the greedy algorithm process, the adaptive optimization algorithm process for the forward-looking container loading and unloading sequence decision can be referred to Figure 4 .

[0101] The embodiments described above are only descriptions of the preferred embodiments of the present invention, and do not limit the scope of the present invention. Without departing from the design spirit of the present invention, various deformations and improvements made by those of ordinary skill in the art to the technical solutions of the present invention shall fall within the protection scope determined by the claims of the present invention.

Claims

1. A prospective and adaptive optimization method for container handling sequence decision-making, characterized in that the steps Including: Calculating the total profit and loss time of candidate operations based on the operation status of containers; Making a locally optimal selection based on the total profit and loss time; Executing the selected operation based on the locally optimal selection; Performing dynamic optimization of the sequence according to the deviation or interruption in real-time operations.

2. The prospective and adaptive optimization method for the container handling sequence decision according to claim 1, characterized in that The expression of the total profit and loss time is as follows: Among them, B s represents the total profit and loss time caused by taking the loading / unloading operation at the container position s as the next operation; represents the current trajectory profit and loss time brought to the empty movement by taking the loading / unloading operation at the container position s as the next operation; represents the current trajectory profit and loss time brought to the heavy movement by taking the loading / unloading operation at the container position s as the next operation; represents the current landing profit and loss time caused by taking the loading / unloading operation at the container position s as the next operation; represents the future trajectory profit and loss time that will inevitably be brought to the subsequent operation process after completing the loading / unloading operation at the container position s; represents the future landing profit and loss time that will inevitably be brought to the subsequent operation process after completing the loading / unloading operation at the container position s.

3. The prospective and adaptive optimization method for container handling sequence decision-making according to claim 2, characterized in that The said and The calculation method includes: Among them, r and s represent the positions of the containers inside the ship's bays, and W D represents the set of positions of the containers unloaded from the ship; W L represents the set of positions of the containers loaded onto the ship; X rs = 1 indicates that the container at s is operated immediately after the container at r; i0 is the lowest ship's bay section, representing the ship's bay section when all the loaded and unloaded containers are emptied; i s is the ship's bay section when loading / unloading the container at position s; represents the empty spreader movement trajectory time from the truck position o to the container position s under the current ship's bay section i s ; represents the waiting time for releasing the previous unloaded container and grabbing the next loaded container at the truck position; represents the empty spreader movement trajectory time from the container position r to the container position s under the current ship's bay section i s ; represents the empty spreader movement trajectory time from the container position r to the truck position o under the current ship's bay section i s ; represents the loaded spreader movement trajectory time from the container position s to the truck position o under the current ship's bay section i s ; represents the loaded spreader movement trajectory time from the container position s to the truck position o at the lowest ship's bay section i0; represents the loaded spreader movement trajectory time from the truck position o to the container position s under the current ship's bay section i s ; represents the loaded quay crane spreader movement trajectory time from the truck position o to the container position s at the lowest ship's bay section i0; represents the empty spreader landing time when grabbing the container at s under the current ship's bay section i s ; represents the loaded spreader landing time when placing the container at s under the current ship's bay section i s ; t L1 represents the shortest landing time when the loaded container needs to be placed and landed in the ship's bay and there is an adjacent container to rely on; t D1 represents the shortest landing time when the unloaded container needs to be grabbed and landed in the ship's bay and the landing guide can be lowered or there is an adjacent container to rely on.

4. The prospective and adaptive optimization method for container handling sequence decision according to claim 3, characterized in that The said and calculation methods include: Among them, γ represents the position of the container in the ship's bay; Y sγ = 1 indicates that the container located at γ is operated after the container located at s; i s is the ship's bay section before loading / unloading the container located at position s, and i s’ is the ship's bay section after loading / unloading the container located at position s; represents the time of the heavy load movement trajectory of the spreader from the gantry position o to the container position r under the ship's bay section after the container located at s is loaded / unloaded; represents the time of the heavy load movement trajectory of the spreader from the gantry position o to the container position r under the ship's bay section before the container located at s is loaded / unloaded; represents the landing time of the heavy load spreader when placing the container located at r under the ship's bay section after the container located at s is loaded / unloaded; represents the landing time of the heavy load spreader when placing the container located at r under the ship's bay section before the container located at s is loaded / unloaded; represents the landing time of the empty load spreader when gripping the container located at r under the ship's bay section after the container located at s is loaded / unloaded; represents the landing time of the empty load spreader when gripping the container located at r under the ship's bay section before the container located at s is loaded / unloaded.

5. The prospective and adaptive optimization method for container handling sequence decision-making according to claim 1, characterized in that The steps for making the locally optimal selection include: based on a heuristic greedy algorithm, determining the loading or unloading task of the next operation through local decisions at each step, and finally forming a complete container loading and unloading sequence plan.

6. A prospective and adaptive optimization system for container handling sequence decision-making, the system being used to implement the method according to any one of claims 1-5, characterized in that, Including: A calculation module, a selection module, an execution module, and an optimization module; The calculation module is used to calculate the total profit and loss time of candidate operations based on the operation status of containers; The selection module is used to make a locally optimal selection based on the total profit and loss time; The execution module is used to execute the selected operation based on the locally optimal selection; The optimization module is used to perform dynamic optimization of the sequence according to the deviation or interruption in real-time operations.

7. The prospective and adaptive optimization system for container handling sequence decision-making according to claim 6, characterized in that The expression of the total profit and loss time is as follows: Among them, B s represents the total profit and loss time caused by taking the loading / unloading operation at container position s as the next operation; represents the current trajectory profit and loss time brought to the empty movement by taking the loading / unloading operation at container position s as the next operation; represents the current trajectory profit and loss time brought to the heavy movement by taking the loading / unloading operation at container position s as the next operation; represents the current positioning profit and loss time caused by taking the loading / unloading operation at container position s as the next operation; represents the future trajectory profit and loss time that will inevitably be brought to the subsequent operation process after completing the loading / unloading operation at container position s; represents the future positioning profit and loss time that will inevitably be brought to the subsequent operation process after completing the loading / unloading operation at container position s.

8. The prospective and adaptive optimization system for container handling sequence decision-making according to claim 6, characterized in that The workflow of the selection module includes: based on a heuristic greedy algorithm, determining the loading or unloading task of the next operation through local decisions at each step, and finally forming a complete container loading and unloading sequence plan.

Images