Freight container transport dispatch management system

Abstract

The application discloses a freight container transportation scheduling management system and relates to the technical field of transportation scheduling.The freight container transportation scheduling management system can realize the precision, collaboration and dynamic optimization of container scheduling, improve the utilization rate of wharf storage space and the resource adaptation efficiency, reduce operation delay and resource waste, guarantee the efficient and stable operation of the whole transportation scheduling process, and solve the problems of fuzzy priority determination, lack of pertinence in storage space allocation, low operation collaboration efficiency and inability to timely correct instruction execution deviation in traditional scheduling.

Description

Technical Field

[0001] This invention relates to the field of transportation scheduling technology, specifically a freight container transportation scheduling and management system. Background Technology

[0002] In actual terminal operations, dispatching systems often face complex conditions such as frequent arrivals of multiple vessels, mixed high-value cargo, dynamic malfunctions of loading and unloading equipment, and fluctuating container pickup demand. Existing dispatching schemes reveal the following shortcomings: Although existing schemes have priority calculation methods based on multi-dimensional data, their weight sets are mostly statically set or simply weighted, lacking the ability to adaptively match different operational scenarios, leading to delays in loading or picking up critical containers. Furthermore, priority determination, cargo space allocation, and gantry crane configuration are mostly independent decision-making processes, lacking a collaborative optimization mechanism based on historically similar scenarios.

[0003] Therefore, there is an urgent need for an intelligent freight container scheduling system that can quantify container priorities in multiple dimensions, accurately match collaborative decision-making, and respond to and correct instruction deviations in real time, in order to solve the above-mentioned technical bottlenecks and improve the overall scheduling performance of terminals in complex operating environments. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides a freight container transportation scheduling and management system that solves the problems of ambiguous priority determination, lack of targeted allocation of cargo locations, low efficiency of operational collaboration, and inability to correct instruction execution deviations in a timely manner in traditional scheduling.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a freight container transportation scheduling and management system, comprising: a priority determination module, used to match a weight set from a weighted association database based on the ship departure timeliness, container attribute urgency and arrival container demand, and determine the priority of freight containers.

[0006] The cargo space allocation module is used to allocate cargo spaces to freight containers of different priorities by comparing the coordinate distance difference between available cargo spaces and the frequency of cargo space usage, based on priority and the coordinates of available cargo spaces.

[0007] The transportation scheduling module is used to perform a weighted summation of priority, compatibility between cargo location and gantry crane, and compatibility between gantry crane and work point to obtain the operational coordination compatibility of each freight container. Then, based on the operational coordination compatibility, a three-dimensional tensor scheduling space is constructed, and collaborative operation instructions are generated through lexicographical slot competition and real-time preemption.

[0008] The operation instruction adjustment module is used to analyze time deviation, gantry crane adaptation deviation, and step completion deviation when scheduling transportation based on collaborative operation instructions. It adjusts the historical operation deviation vector in the case library by matching the Euclidean distance to obtain the adjustment judgment value. If the adjustment judgment value is greater than the set judgment threshold, the collaborative operation instruction is adjusted.

[0009] Compared with existing technologies, the present invention has the following beneficial effects: the present invention accurately matches the urgent needs of containers through a priority determination module, optimizes the configuration of cargo locations based on priority and coordinates through a cargo location allocation module, generates efficient collaborative operation instructions by combining multi-dimensional adaptability through a transportation scheduling module, and responds to execution deviations in real time through an operation instruction adjustment module. This enables precise, collaborative, and dynamic optimization of container scheduling, improves terminal cargo location utilization and resource adaptation efficiency, reduces operation delays and resource waste, and ensures the efficient and stable operation of the entire transportation scheduling process. It can solve the problems of ambiguous priority determination, lack of targeted cargo location allocation, low operation collaboration efficiency, and inability to correct instruction execution deviations in a timely manner in traditional scheduling. Attached Figure Description

[0010] Figure 1 This is a schematic diagram of the module connections of the freight container transportation scheduling and management system of the present invention;

[0011] Figure 2 This is a flowchart illustrating the process of determining the priority of freight containers in the freight container transportation scheduling and management system of the present invention.

[0012] Figure 3 This is a flowchart illustrating how the freight container transportation scheduling and management system of the present invention obtains adjustment judgment values. Detailed Implementation

[0013] The present invention will now be described in detail with reference to the accompanying drawings and embodiments. Please refer to the accompanying drawings. Figure 1 This invention provides a technical solution: a freight container transportation scheduling and management system, including a priority determination module, a cargo location allocation module, a transportation scheduling module, and an operation instruction adjustment module. The priority determination module is connected to the cargo location allocation module, the cargo location allocation module is connected to the transportation scheduling module, and the transportation scheduling module is connected to the operation instruction adjustment module. The priority determination module is used to determine the priority of freight containers by matching a weight set from a weighted association database based on the ship's departure timeliness, the urgency of container attributes, and the demand for arriving containers.

[0014] The closer a ship's departure time, the more priority its containers need to be given to avoid being left unattended. The properties of the goods inside the containers (such as perishable goods, dangerous goods, and emergency supplies) dictate that they cannot remain in place for extended periods. Containers arriving at the port should also be prioritized if the recipient's pickup time is urgent. However, the relative importance of these three factors is not fixed. Using a fixed-weighted calculation to determine priority cannot adapt to the dynamic changes in the importance of each factor under different operational scenarios, leading to distorted priority determination. Therefore, a strategy based on Euclidean distance matching of historical similar scenarios' weight sets is adopted: by matching the three standardized parameters of the current container with parameters in historical records, the scenario with the most similar historical operational environment is found, and the validated weight set for that scenario is reused to dynamically and adaptively determine the priority of the current container. Figure 2 As shown, the specific process is as follows: the ship departure time, container attribute urgency and arrival container demand of each freight container are standardized to obtain a standardized parameter set.

[0015] It should be noted that the process of obtaining the vessel departure time is as follows: the planned departure time of the vessel and the current time of the dispatch system are collected, and the time between the planned departure time of the vessel and the current time of the dispatch system is used as the departure time window.

[0016] The departure timeliness of a vessel is determined by the departure time window: If the departure time window is no more than 24 hours, the departure timeliness is 100. If the departure time window is greater than 24 hours but no more than 72 hours, the departure timeliness is 70. If the departure time window is greater than 72 hours but no more than 120 hours, the departure timeliness is 40. If the departure time window is greater than 120 hours, the departure timeliness is 10. The departure timeliness measures the urgency of the vessel to which the container belongs; the higher the value, the closer the vessel is to departure, and the higher the priority of handling the container.

[0017] The process of obtaining the urgency of container attributes is as follows: collect cargo attributes, including cargo type, preservation or storage time limit, and special transportation requirements.

[0018] Scoring is done based on the acquired attributes. For example, if the goods are emergency supplies, the score is 40 points; if the preservation / storage time is no more than 24 hours, the score is 30 points; and if there is a special transportation requirement, the score is 10 points.

[0019] The sum of the scores yields the container attribute urgency level, which reflects the urgent handling needs of the goods inside the container. The higher the value, the more priority the goods need to be transported (such as perishable goods, emergency supplies, etc.).

[0020] The process of obtaining the demand for arriving containers is as follows: collect the estimated arrival time of the containers and the scheduled pickup time of the container pick-up person, and calculate the time difference between the two.

[0021] The demand index for arriving containers is determined by time difference: If the time difference is no more than 24 hours, the demand index is 100. If the time difference is greater than 24 hours but no more than 72 hours, the demand index is 70. If the time difference is greater than 72 hours but no more than 120 hours, the demand index is 40. If the time difference is greater than 120 hours, the demand index is 10. The container demand index measures the urgency with which the person picking up the container needs to collect it; a higher value indicates that the scheduled pickup time is approaching and the container needs to be placed as soon as possible.

[0022] Retrieve the standardized historical parameter set stored in the weighted association database, and calculate the similarity between the standardized parameter set and each standardized historical parameter set in the weighted association database to obtain the corresponding Euclidean distance. The smaller the distance, the more similar the urgency characteristics of the current parameter combination and the historical parameter combination are.

[0023] After calculating the distance for all standardized historical parameter sets, a list of associations between the standardized historical parameter sets and Euclidean distances is generated, recording the Euclidean distance for each historical scene.

[0024] Sort all Euclidean distances in the association list numerically, determine the standardized historical parameter set corresponding to the smallest Euclidean distance, and extract the weight set corresponding to the standardized historical parameter set from the weight association database.

[0025] It should be noted that if there are two or more standardized historical parameter sets with equal Euclidean distances and both being minimum values, one of them is randomly selected to ensure that the weight set is unique and the sum of all parameters in the weight set is 1.

[0026] The priority is obtained by weighted summation of the standardized parameter set and the weight set.

[0027] The weighted association database can be updated with historical scheduling records, and Euclidean distance matching can automatically adapt to new scheduling scenarios without frequent adjustments to the judgment rules, thus improving the system's adaptability to changes in dock scheduling scenarios.

[0028] This embodiment uses a priority calculation method based on historical similarity scenario matching. This eliminates the need for frequent manual adjustments to weight parameters, ensuring that the priority ranking results better reflect the current operational urgency. This effectively reduces the probability of high-priority containers being delayed in loading or picking up due to weight mismatch. Furthermore, since Euclidean distance calculation and weighted summation are both lightweight operations, this scheme can quickly output reliable priority values ​​without increasing the burden on the scheduling system.

[0029] It should be noted that the construction method of the weighted association database is as follows: For each historical scheduling record, three indicators are collected: container scheduling operation delay time, container pickup task completion rate, and terminal gantry crane occupancy efficiency. These three indicators are normalized to ensure they are of the same order of magnitude. Then, the entropy weight method is used to calculate the objective weight of each indicator: the entropy weight method determines the weight based on the dispersion of the data for each indicator; the greater the data dispersion, the smaller the information entropy, and the greater the weight. The three normalized indicators are multiplied by their respective entropy weights, and then summed to obtain the priority quantification value of that record.

[0030] The recorded ship departure timeliness, container attribute urgency, and arrival container demand were standardized. Then, Pearson correlation coefficients were calculated using these coefficients and the priority quantification values ​​obtained in the first step, resulting in three correlation coefficients. The Pearson correlation coefficient reflects the direction and degree of linear correlation between two variables.

[0031] Take the absolute values ​​of the three correlation coefficients mentioned above, and then sum them to obtain the total absolute value. Divide the absolute value of each correlation coefficient by the total absolute value to obtain three normalized weight values, corresponding to the ship departure timeliness, container attribute urgency, and arrival container demand, respectively. These three weight values ​​constitute the weight set corresponding to this record.

[0032] The standardized ship departure time, container attribute urgency, and arrival container demand are recorded and associated with the weight set obtained in step three, forming a record in the weighted association database. The above steps are repeated to process all historical scheduling records, thus completing the database construction.

[0033] In addition, the weighted association database adopts a sliding window update mechanism: based on the historical scheduling records of the most recent 90 days, the priority quantification value and Pearson correlation coefficient are recalculated for the records within the window every 7 days, and the weight set corresponding to each record is updated; records older than 90 days are automatically removed from the database. When the number of newly added scheduling records reaches 100, an incremental update is triggered, which only calculates the weight set for the newly added records and stores them in association, without reconstructing the entire database.

[0034] It should be noted that container dispatching operation delay time refers to the absolute value of the difference between the actual total completion time and the planned completion time of historical operations. Container pickup completion rate refers to the ratio of the number of containers actually picked up by the pickup personnel in historical operations to the planned number of containers to be picked up. Terminal gantry crane occupancy efficiency refers to the ratio of the actual usage time of the gantry cranes used in historical operations to the planned allocated time.

[0035] Considering that the three data points of container scheduling delay time, container pick-up task completion rate, and terminal gantry crane utilization efficiency are inconsistent in terms of their respective advantages and disadvantages, they need to be normalized according to the principle of positive uniformity.

[0036] For example, the initial weight set for weighted summation is: container dispatching operation delay time 0.4, container pickup task completion rate 0.35, and terminal gantry crane utilization efficiency 0.25. This weight set can be optimized and updated through regression analysis of historical dispatching data from the past 3 months.

[0037] After standardizing the ship departure timeliness, container attribute urgency, and arrival container demand for each historical scheduling record, Pearson correlation coefficients were calculated with the priority quantification values ​​to obtain three correlation coefficients.

[0038] The absolute values ​​of the three correlation coefficients corresponding to each historical scheduling record are processed and normalized to obtain a weight set.

[0039] It should be noted that the normalization process is as follows: first, the sum of the absolute values ​​of the three correlation coefficients is calculated as the total correlation degree.

[0040] Then, the ratio of the absolute value of each of the three correlation coefficients to the total correlation is calculated and used as the weight of the corresponding parameter.

[0041] By associating the standardized ship departure time, container attribute urgency, and arrival container demand corresponding to each historical scheduling record with a weight set, a weighted association database is obtained.

[0042] The cargo space allocation module is used to allocate cargo spaces to freight containers of different priorities by comparing the coordinate distance difference between available cargo spaces and the frequency of cargo space use, based on priority and the coordinates of available cargo spaces. Specifically, it constructs a three-dimensional coordinate system centered on the quay crane loading and unloading point to determine the coordinates of each freight container and the coordinates of available cargo spaces.

[0043] For example, the X-axis of the three-dimensional coordinate system is along the quayline, the Y-axis is perpendicular to the quayline extending into the storage yard, and the Z-axis is perpendicular to the ground height.

[0044] The 3D coordinates of all available cargo locations are extracted through the terminal cargo management system. The geometric center point of each cargo location is used as the coordinate collection point, such as the center point coordinates of a 20-foot standard cargo location. For outbound containers, the center point coordinates of their current temporary storage point are used, and for inbound containers, the center point coordinates of their initial target area to be temporarily stored after unloading are used. All of these are obtained in real time through the scheduling system.

[0045] To ensure that high-urgency containers are given priority access to quality cargo spaces, the coordinates of each freight container are sorted from highest to lowest priority. The coordinate distance between the coordinates of the highest priority freight container and the coordinates of each available cargo space is calculated. Based on the coordinate distance values, the available cargo spaces are sorted in ascending order to obtain a candidate list for cargo space allocation.

[0046] The frequency of use of a storage location refers to the total number of times that an available storage location has been allocated and completed within a preset historical period T (e.g., the past 30 days), divided by the total number of working days in that period to obtain the average daily usage frequency.

[0047] Considering that there may be two or more vacant cargo spaces with similar or equal coordinate distances in the cargo space allocation candidate list, in order to determine which cargo space to use, it is necessary to find the coordinate distance between the first and second vacant cargo spaces after obtaining the cargo space allocation candidate list, and then calculate the difference between the coordinate distance between the second and first vacant cargo spaces in the cargo space allocation candidate list. If the number of vacant cargo spaces is equal to 1, then the first vacant cargo space is directly allocated to the freight container with the highest priority.

[0048] If the difference is greater than the set distance threshold, the nearest available cargo space will be allocated to the freight container with the highest priority.

[0049] If the difference is not greater than the set distance threshold, the usage frequency of the second available cargo space is compared with that of the first available cargo space, and the available cargo space with the highest usage frequency is assigned to the freight container corresponding to the highest priority.

[0050] After allocating cargo space to the highest priority container, repeat the above steps to re-lock a new highest priority container from the remaining unallocated containers, and continue to calculate distances, compare thresholds, and allocate cargo space until all unallocated containers have been allocated cargo space or there are no available cargo spaces.

[0051] It should be noted that the distance threshold is set at 5 meters. This value is determined based on the operating radius of a 20-foot standard container and the loading and unloading efficiency of the quay crane. If the terminal mainly uses 40-foot containers, it can be adjusted to 8 meters.

[0052] The transportation scheduling module is used to perform a weighted summation of priority, compatibility between cargo location and gantry crane, and compatibility between gantry crane and work point to obtain the operational coordination compatibility of each freight container. Then, based on the operational coordination compatibility, a three-dimensional tensor scheduling space is constructed, and collaborative operation instructions are generated through lexicographical slot competition and real-time preemption.

[0053] The process of determining the operational coordination and adaptability of each freight container is as follows: obtain historical completed operation records stored in the system, and calculate the deviation between the actual operation time and the planned operation time of the historical completed operation records.

[0054] It should be noted that the formula for calculating the deviation is: ,in, For deviation degree, The actual time spent on the task. The planned time. If If the calculated value is less than 0, then it is taken as 0. The closer the value is to 1, the better the actual collaboration effect of the historical operation matches the plan, and the better the collaboration. The closer the value is to 0, the worse the collaboration effect.

[0055] Based on all historical completed operation records, Pearson correlation coefficients were calculated between priority and deviation, fit and deviation between the cargo location and the gantry crane, and fit and deviation between the gantry crane and the work point. The correlation coefficients corresponding to priority, fit and deviation between the cargo location and the gantry crane, and fit and deviation between the gantry crane and the work point were obtained.

[0056] Before obtaining the compatibility between the storage location and the gantry crane, the process also includes the step of determining the compatible gantry crane: for each vacant storage location or assigned storage location, select one gantry crane from multiple available gantry cranes that minimizes the reachability distance between the gantry crane and the storage location and has a gantry crane compatibility greater than or equal to 1 as the compatible gantry crane.

[0057] The process of obtaining the compatibility between the cargo location and the gantry crane is as follows: calculate the reachable distance of the gantry crane cargo location based on the coordinates of the cargo location and the coordinates of the compatible gantry crane.

[0058] Calculate the ratio of the maximum load capacity of the compatible gantry crane to the actual weight of the container and the ratio of the working space dimensions of the compatible gantry crane to the actual dimensions of the container. Multiply the two ratios to obtain the gantry crane compatibility.

[0059] The compatibility between the gantry crane and the gantry crane is obtained by multiplying the gantry crane's reachability and compatibility after standardization.

[0060] The compatibility between the cargo location and the gantry crane is used to quantify the spatial matching and physical compatibility between the target cargo location and the gantry crane. The higher the value, the stronger the compatibility.

[0061] The process of obtaining the compatibility between the gantry crane and the work point is as follows: calculate the reachable distance of the gantry crane work point based on the coordinates of the work point and the coordinates of the compatible gantry crane.

[0062] Calculate the ratio of the current busy time of the adapted gantry crane to its rated working cycle, and use the difference between 1 and the ratio as the availability of the adapted gantry crane. The current busy time of the gantry crane refers to the sum of the remaining time of the work currently being performed by the gantry crane and the planned time of the scheduled but not yet started work, which is read in real time from the gantry crane control system.

[0063] The compatibility between the gantry crane and the work point is obtained by multiplying the standardized reach distance of the gantry crane work point and the availability of the suitable gantry crane.

[0064] The compatibility between the gantry crane and the work site is used to quantify the spatial matching between the gantry crane and the target work site, such as the outbound quay crane and the inbound container pick-up area, as well as the gantry crane's responsiveness. The higher the value, the stronger the compatibility.

[0065] The absolute values ​​of the correlation coefficients are processed and normalized to obtain the priority, the compatibility between the cargo location and the gantry crane, and the compatibility between the gantry crane and the work point, respectively, and their corresponding weight factors.

[0066] The operational coordination and adaptation degree is obtained by multiplying the priority, the compatibility between the cargo location and the gantry crane, and the compatibility between the gantry crane and the work point by their respective weighting factors and summing them.

[0067] Given the tight coupling of time, space, and equipment in port operations, independent scheduling would lead to resource fragmentation and frequent conflicts. Directly allocating resources based on priority could easily result in local optima while neglecting global resource balance. Therefore, a unified scheduling spatial structure is needed to express when, which equipment, and in which area to use, allowing each available resource combination to become a clearly defined slot, thus transforming the scheduling problem into slot competition and allocation. Simultaneously, due to varying container urgency, high-priority tasks must have an absolute advantage in slot competition, but simply covering all low-priority tasks cannot be done crudely, otherwise it would cause a large number of reschedulings. Therefore, a lexicographical comparison rule is introduced, comparing tasks in descending order of their impact on scheduling effectiveness. This ensures the priority optimization of key indicators while avoiding the difficulty in determining weights in multi-objective weighted calculations. Furthermore, dynamically arriving tasks and unpredictable operation durations require the system to handle conflicts instantly. Therefore, an immediate preemption mechanism is added, whereby high-priority tasks deprive low-priority tasks of their already occupied slots, ensuring that urgent tasks are not blocked.

[0068] Specifically, after determining the collaborative adaptability, a three-dimensional tensor scheduling space is constructed based on the operational collaborative adaptability. The process of generating collaborative operation instructions through lexicographical slot competition and real-time preemption is as follows: the current scheduling time window is divided into continuous time slices, all available gantry cranes are used as equipment dimensions, and the yard is divided into several operational sectors as spatial dimensions. A three-dimensional tensor scheduling space is constructed with time slices, gantry crane indexes, and operational sector indexes as three dimensions. Each unit in the three-dimensional tensor scheduling space is an idle schedulable slot.

[0069] For each cargo container to be scheduled, the operating sector to which it belongs is determined based on the time slice range allowed by the ship's departure time limit, the coordinates of the cargo location already assigned to the cargo container, and the available equipment channel is determined based on the number of the gantry crane that is compatible with the cargo container. The available slots that simultaneously meet the time slice range, operating sector, and equipment channel requirements are selected from the three-dimensional tensor scheduling space as the candidate slot set for the cargo container.

[0070] The optimal slot for the current candidate slots of the freight container is selected using a lexicographical order rule. The comparison order is as follows: First, calculate the congestion level of the time slice containing the slot (number of occupied slots divided by the total number of slots). Then, query the current task chain length (number of unexecuted instructions) of the gantry crane corresponding to the slot. Next, determine the time period type based on the start time of the time slice (e.g., assign 3 for peak hours, 2 for off-peak hours, and 1 for off-peak hours). Finally, retrieve the pre-calculated operational coordination suitability of the container. Following the lexicographical order rule of lower congestion level, shorter task chain length, lower time period type value, and higher operational coordination suitability, all candidate slots are compared sequentially: first, the congestion level is compared, with the lowest congestion level winning. If the congestion levels are the same, the task chain length is compared, with the shorter task chain winning. If still the same, the time period type is compared, with the smaller value winning. If still the same, the operational coordination suitability is compared, with the larger value winning. Finally, the optimal slot is selected and allocated to the container.

[0071] When multiple freight containers have the same available slot in their candidate slots, the freight container with higher priority is deprived of the same slot that has been assigned to the freight container with lower priority. The freight container whose slot has been deprived re-enters the waiting queue and is moved back to the next time slice.

[0072] Once all the freight containers to be dispatched have been assigned a unique slot, each slot is mapped to a collaborative operation instruction. Each collaborative operation instruction includes at least the operation start time, the assigned gantry crane number, and the sequence of operation path points. The collaborative operation instruction is then sent to the gantry crane control system and the on-site operation terminal.

[0073] A time slice is a fixed-length period of time that continuously divides the current scheduling time window, for example, every 10 minutes. The numerical value represents the sequence number of the time slice on the timeline; the smaller the sequence number, the earlier the time. Its length determines the time resolution of the scheduling; the shorter the length, the more refined the scheduling but the greater the computational load.

[0074] A gantry crane index is a unique identifier for all available gantry cranes. A work sector index is a unique number assigned to each work area after the yard is divided into several work zones (e.g., Zone A, Zone B).

[0075] An idle schedulable slot is a specific unit in the three-dimensional tensor scheduling space, represented as (time slice, gantry crane index, work sector index).

[0076] The candidate slot set is a collection of all available slots that meet the constraints of each container to be scheduled, selected from the three-dimensional tensor scheduling space. It is obtained by: determining the allowable time slice range based on the container's ship departure time (e.g., a time slice within 4 hours before departure); determining the operating sector based on the cargo location coordinates; and determining the available equipment channels based on the compatible gantry crane number. The intersection of these three conditions is then used. A quantity of zero indicates no available resources, requiring constraint adjustment or waiting.

[0077] The congestion level of a time slice reflects the ratio of the number of occupied slots within that time slice to the total number of slots in that time slice. It is obtained by: for each time slice, counting the number of occupied slots in all combinations (gantry crane index, work sector index) within that time slice, and dividing that number by the total number of combinations. The value ranges from 0 to 1; the closer to 1, the more congested the time slice, and the more likely subsequent tasks are to conflict. The closer to 0, the less congested the time slice.

[0078] The current task chain length for the gantry crane corresponding to the slot represents the number of work tasks assigned to that gantry crane in the current scheduling plan but not yet executed. It is obtained by scanning the collaborative work instruction queue for instructions assigned to that gantry crane with a start time after the current time. A longer length indicates a heavier workload for the gantry crane, potentially causing long waiting times after new tasks are inserted. A shorter length indicates less idle time for the equipment.

[0079] The time slot type of a docking station is categorized based on the time period it falls within (e.g., morning peak, midday, evening peak, night). This is obtained by combining historical terminal workload statistics, dividing a 24-hour day into several time slots (e.g., 7:00-9:00 is peak, 22:00-6:00 is off-peak), and assigning each time slot to its corresponding time slot type based on its start time. The numerical value is usually represented by a code (e.g., 1 = off-peak, 2 = off-peak, 3 = peak). A higher value indicates more operational interference factors during that time slot (e.g., traffic congestion, staff fatigue), resulting in higher scheduling costs.

[0080] This embodiment employs a three-dimensional tensor scheduling space to unify the modeling of time slices, gantry crane indexes, and work sector indexes. This allows the system to intuitively and completely grasp all available resource slots, solving the problem of fragmented time, equipment, and spatial information in traditional scheduling that hinders collaborative decision-making. Through lexicographical slot competition, the optimal slot is selected for each container from the candidate slot set according to priority. The lexicographical comparison order directly corresponds to the criticality affecting operational efficiency, avoiding the drawbacks of subjective weight setting in weighted summation. This ensures that high-priority containers obtain high-quality slots without sacrificing overall efficiency. The instant preemption mechanism allows high-priority containers to directly preempt lower-priority containers already occupying slots when multiple containers compete for the same slot. The preempted container is then moved back to a later time slice to wait, ensuring that urgent tasks are not blocked and reducing the probability of critical containers being delayed in loading or picking up due to resource conflicts. Overall, this improves the coordination and timeliness of scheduling instructions and reduces resource idle time and task waiting time.

[0081] The operation instruction adjustment module is used to analyze time deviation, gantry crane adaptation deviation, and step completion deviation when scheduling transportation based on collaborative operation instructions. It adjusts the historical operation deviation vector in the case library by matching the Euclidean distance to obtain the adjustment judgment value. If the adjustment judgment value is greater than the set judgment threshold, the collaborative operation instruction is adjusted.

[0082] It should be noted that each historical collaborative operation record in the adjustment case library stores not only the historical operation deviation vector but also a corresponding collaborative operation adjustment instruction. The method for generating collaborative operation adjustment instructions is as follows: When a deviation occurs during the execution of a historical operation and is ultimately completed through manual intervention or system rescheduling, the system records the entire action sequence of the adjustment process, including but not limited to: the reassigned gantry crane number, the modified travel path, the adjusted container operation sequence, and the waiting time threshold. This action sequence is converted into a structured collaborative operation adjustment instruction according to a predefined instruction template (containing four fields: resource ID, action type, timestamp, and coordinate list), and stored in association with the historical operation deviation vector. If multiple identical or similar deviation vectors correspond to different adjustment instructions, the instruction that causes the shortest subsequent delay is taken as the standard instruction.

[0083] like Figure 3 As shown, the process of obtaining the adjustment judgment value is as follows: For the freight container that is executing the collaborative operation instruction, the time deviation, gantry crane adaptation deviation and step completion deviation are collected in real time. The time deviation, gantry crane adaptation deviation and step completion deviation are standardized and combined in a fixed dimension order to obtain the current operation deviation vector.

[0084] The time deviation is the absolute value of the difference between the current cumulative actual time spent on the operation and the preset cumulative time spent on the command. The gantry crane adaptation deviation is the ratio of the difference between the current actual load of the operating gantry crane and the preset load on the command, calculated by comparing the absolute value of the difference between the actual load and the preset load with the preset load. The step completion deviation is the difference between the number of operation steps that should be completed and the number of steps that have actually been completed, where the number of operation steps that should be completed is the preset sequence of operation steps in the collaborative operation command.

[0085] The current job deviation vector is compared with the historical job deviation vectors of each historical collaborative job record in the adjustment case library using Euclidean distance calculation. The distance value is then used as the similarity representation value; the smaller the distance, the higher the similarity.

[0086] Sort the historical collaborative work records in ascending order of similarity scores, select the top N records by similarity score, and obtain their corresponding adjusted label values. The adjusted label values ​​are pre-defined level values ​​based on the degree of influence of the historical collaborative work records.

[0087] Selecting the top N historical collaborative work records with the highest similarity values ​​ensures that the historical collaborative work records used to calculate the adjustment judgment value are highly consistent with the current deviation characteristics. Compared with randomly selecting historical collaborative work records, the adjustment labels of similar historical collaborative work records are more valuable for reference.

[0088] For example, N is usually 5-10. From the selected N records, the corresponding adjustment label values ​​are extracted. The adjustment label values ​​are 1-4, which correspond to no adjustment, slight adjustment, moderate adjustment, and urgent adjustment, respectively.

[0089] By statistically analyzing the proportion of each adjustment label value in N historical collaborative work records, we can reflect the probability distribution of different adjustment necessities under similar historical deviation scenarios. For example, if 5 out of 7 historical collaborative work records require moderate adjustment, 1 requires slight adjustment, and 1 requires urgent adjustment, the proportion will objectively reflect that moderate adjustment is the mainstream demand.

[0090] The adjustment judgment value is obtained by multiplying each percentage by the corresponding adjustment label value and summing the results. The adjustment judgment value ranges from 1 to 4. The larger the value, the more similar the current operation deviation is to the deviation characteristics that require high-level adjustment in history, and the higher the necessity of adjusting the collaborative operation instruction.

[0091] It should be noted that the process of obtaining the adjustment tag value corresponding to the historical collaborative operation record is as follows: for each historical collaborative operation record, the duration of the collaborative chain effect after the execution of each historical operation and the frequency of gantry crane rescheduling are standardized and then assembled into an impact vector in a fixed dimension order to ensure that the impact vector can objectively reflect the actual severity of the impact caused by the operation deviation.

[0092] The duration of the synergistic chain effect refers to the total delay time of subsequent related operations caused by initial deviations in historical operations, such as time deviations or resource adaptation deviations. The frequency of gantry crane rescheduling refers to the actual number of times the gantry crane equipment has been temporarily adjusted or replaced to correct the historical operation deviation.

[0093] The elbow rule is used to determine the value of K, which serves as the number of clusters. The influence vectors are then clustered into K clusters using the K-means clustering algorithm. The mean influence vector of each cluster is calculated; specifically, K historical influence vectors are randomly selected as the initial cluster centroids.

[0094] Calculate the Euclidean distance from each historical influence vector to the K initial cluster centroids, and assign the historical influence vector to the group containing the nearest centroid.

[0095] The new centroid of each group is recalculated based on the mean of all historical influence vectors within that group.

[0096] The process of iterating through grouping and updating centroids is repeated until the set number of iterations is reached, ultimately forming K clusters.

[0097] The K-means algorithm can automatically adapt to the distribution of consequence data in different terminals and business scenarios: if a terminal adds an automated gantry crane, the frequency of gantry crane rescheduling may decrease, the historical impact vector of newly added historical records will present new features, and the algorithm will automatically adjust the cluster group boundaries when re-clustering without the need for manual model reconstruction.

[0098] Calculate the magnitude of the mean influence vector of each cluster group, sort the cluster groups in ascending order of magnitude, use the sequence number as the adjustment label value corresponding to each historical collaborative work record in the cluster group, bind the influence vector of each historical collaborative work record with the corresponding adjustment label value, and update the adjustment case library.

[0099] For example, the sorted cluster group number is directly used as the adjustment label value for all historical collaborative work records within that group. The mapping rule is objective and unique: the cluster group with the least impact is sorted as 1 and has an adjustment label value of 1.

[0100] For clusters less affected, sort by 2 and adjust the label value to 2.

[0101] For clusters with moderate impact, sort by 3 → adjust label value to 3.

[0102] The cluster group most affected was ranked 4 and its label value was adjusted to 4.

[0103] The label value directly corresponds to the chain reaction and resource rescheduling consequences caused by operational deviations. The more severe the impact of a cluster, the larger the label value, which can truly reflect the adjustment needs of historical deviations and avoid the label from being out of touch with the actual impact.

[0104] If the adjustment judgment value is greater than the set judgment threshold, the process of adjusting the collaborative operation instruction is as follows: retrieve the adjustment case library, find the historical record that best matches the current operation deviation through similarity calculation, and take the collaborative operation adjustment instruction corresponding to the historical collaborative operation record with the smallest similarity characterization value as the new collaborative operation instruction.

[0105] For example, the judgment threshold ranges from 1.8 to 2.5, preferably 2.0. The judgment threshold can be dynamically adjusted according to the average daily workload of the terminal. When the workload is greater than 500 boxes / day, the threshold is 2.2, and when it is less than 300 boxes / day, the threshold is 1.9.

[0106] The selected optimal historical cases have the highest similarity to the operational scenarios, and their adjustment logic has been verified to be effective in historical scenarios. The historical adjustment instructions with the lowest similarity can be directly reused without redesigning the adjustment plan. This is especially suitable for scenarios with urgent adjustment needs, and avoids further deterioration of operational deviations due to decision-making delays.

[0107] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, in the form of a computer program product.

[0108] Those skilled in the art will recognize that the modules and algorithm steps of the various examples 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 implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0109] In addition, the functional modules in the various embodiments of this application can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module.

[0110] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0111] Finally, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A freight container transportation dispatch and management system, characterized in that, include: The priority determination module is used to match a set of weights from a weighted association database based on the ship's departure timeliness, the urgency of container attributes, and the demand for arriving containers to determine the priority of freight containers. The cargo space allocation module is used to allocate cargo spaces to freight containers of different priorities by comparing the coordinate distance difference between available cargo spaces and the frequency of cargo space usage, based on priority and the coordinates of available cargo spaces. The transportation scheduling module is used to perform a weighted summation of priority, compatibility between cargo location and gantry crane, and compatibility between gantry crane and work point to obtain the operational coordination compatibility of each freight container. Then, a three-dimensional tensor scheduling space is constructed based on the operational coordination compatibility, and collaborative operation instructions are generated through lexicographical slot competition and real-time preemption. The operation instruction adjustment module is used to analyze time deviation, gantry crane adaptation deviation, and step completion deviation when scheduling transportation based on collaborative operation instructions. It adjusts the historical operation deviation vector in the case library by matching the Euclidean distance to obtain the adjustment judgment value. If the adjustment judgment value is greater than the set judgment threshold, the collaborative operation instruction is adjusted.

2. The freight container transportation dispatch and management system according to claim 1, characterized in that, The process of prioritizing freight containers is as follows: For each freight container, the ship departure time, container attribute urgency, and arrival container demand are standardized to obtain a standardized parameter set; The similarity between the standardized parameter set and each standardized historical parameter set in the weight association database is calculated to obtain the corresponding Euclidean distance. Determine the standardized historical parameter set corresponding to the minimum Euclidean distance. If multiple sets exist, select one and extract the corresponding bound weight set from the weight association database. The priority is obtained by weighted summation of the standardized parameter set and the weight set.

3. The freight container transportation dispatch and management system according to claim 2, characterized in that, The construction process of the weighted association database is as follows: After normalizing the container scheduling operation delay time, container pickup task completion rate and terminal gantry crane occupancy efficiency corresponding to each historical scheduling record, a weighted sum is performed to obtain the priority quantification value. After standardizing the ship departure time, container attribute urgency and arrival container demand for each historical scheduling record, Pearson correlation coefficients were calculated with the priority quantification values ​​to obtain three correlation coefficients. The absolute values ​​of the three correlation coefficients corresponding to each historical scheduling record are processed and normalized to obtain a weight set; By associating the standardized ship departure time, container attribute urgency, and arrival container demand corresponding to each historical scheduling record with a weight set, a weighted association database is obtained.

4. The freight container transportation dispatch and management system according to claim 1, characterized in that, The process of allocating cargo space to freight containers of different priorities is as follows: A three-dimensional coordinate system is constructed with the quay crane loading and unloading point as the center to determine the coordinates of each freight container and the coordinates of available cargo space. The freight containers are sorted according to priority. The coordinate distance between the coordinates of the highest priority freight container and the coordinates of each available cargo space is calculated. Based on the coordinate distance, the available cargo spaces are sorted in ascending order to obtain a candidate list of cargo space allocation. If there is only 1 available cargo space, then the only available cargo space will be allocated to the freight container with the highest priority. If the number of available storage locations is greater than or equal to 2, then calculate the difference between the coordinate distance value of the second available storage location in the storage location allocation candidate list and the coordinate distance value of the first available storage location. If the difference is greater than the set distance threshold, the nearest available cargo space will be allocated to the freight container with the highest priority. If the difference is not greater than the set distance threshold, the usage frequency of the second available cargo space is compared with that of the first available cargo space, and the available cargo space with the highest usage frequency is assigned to the freight container corresponding to the highest priority.

5. A freight container transportation dispatching and management system according to claim 1, characterized in that, The process of obtaining the operational coordination and compatibility of each freight container is as follows: Calculate the deviation between the actual time taken and the planned time taken for completed tasks in the historical task records; Based on all historical completed operation records, Pearson correlation coefficients were calculated between deviation degree and priority, fit between cargo location and gantry crane, and fit between gantry crane and operation point, respectively, to obtain the correlation coefficients corresponding to priority, fit between cargo location and gantry crane, and fit between gantry crane and operation point. The absolute values ​​of the correlation coefficients are processed and normalized to obtain the priority, the matching degree between the cargo location and the gantry crane, and the matching degree between the gantry crane and the work point, respectively, and their corresponding weight factors. The operational coordination and adaptation degree is obtained by multiplying the priority, the compatibility between the cargo location and the gantry crane, and the compatibility between the gantry crane and the work point by their respective weighting factors and summing them.

6. A freight container transportation dispatching and management system according to claim 5, characterized in that, The process of obtaining the compatibility between the cargo location and the gantry crane is as follows: Calculate the reachable distance of the gantry crane's cargo location based on the cargo location coordinates and the coordinates of the adapted gantry crane; Calculate the ratio of the maximum load capacity of the compatible gantry crane to the actual weight of the container and the ratio of the working space dimensions of the compatible gantry crane to the actual dimensions of the container. Multiply the two ratios to obtain the gantry crane compatibility. The compatibility between the gantry crane and the gantry crane is obtained by multiplying the gantry crane's reachability and compatibility after standardization. The process of obtaining the compatibility between the gantry crane and the work site is as follows: Calculate the reachable distance of the gantry crane's working point based on the coordinates of the working point and the coordinates of the adapted gantry crane; Calculate the ratio of the current busy time of the adapted gantry crane to the rated working cycle, and take the difference between 1 and the ratio as the availability of the adapted gantry crane; The compatibility between the gantry crane and the work point is obtained by multiplying the standardized reach distance of the gantry crane work point and the availability of the suitable gantry crane.

7. A freight container transportation dispatching and management system according to claim 1, characterized in that, The process of determining collaborative task instructions based on task collaboration adaptability is as follows: The current scheduling time window is divided into continuous time slices, all available gantry cranes are used as equipment dimensions, and the yard is divided into several operating sectors as spatial dimensions to construct a three-dimensional tensor scheduling space. For each cargo container to be scheduled, the operating sector to which it belongs is determined based on the time slice range allowed by the ship's departure time limit, the coordinates of the allocated cargo position, and the available equipment channel is determined based on the matching gantry crane number. The available slots that simultaneously meet the time slice range, operating sector, and equipment channel requirements are selected from the three-dimensional tensor scheduling space as the candidate slot set. The optimal slot is selected from the candidate slots of the current freight container using a lexicographical order rule. The comparison order of the lexicographical order rule is as follows: the first comparison item is the congestion of the time slice in which the slot is located; the second comparison item is the current task chain length of the gantry crane corresponding to the slot; the third comparison item is the time period type of the time slice in which the slot is located; and the fourth comparison item is the operational coordination and adaptability of the freight container. When multiple freight containers have the same available slot in their candidate slots, the freight container with higher priority is deprived of the same slot that has been allocated to the freight container with lower priority. The freight container whose slot has been deprived re-enters the waiting queue and is moved back to the next time slice. Once all the freight containers awaiting scheduling have been assigned a unique slot, each slot is mapped to a collaborative operation instruction.

8. A freight container transportation dispatching and management system according to claim 1, characterized in that, The process of obtaining the adjustment judgment value is as follows: The time deviation, gantry crane adaptation deviation, and step completion deviation are standardized to obtain the current operation deviation vector. The current job deviation vector is compared with the historical job deviation vectors of each historical collaborative job record in the adjustment case library using Euclidean distance calculation to obtain the distance value, and the distance value is used as the similarity representation value. Sort the similarity values ​​in ascending order, select the top N historical collaborative work records with the highest similarity values, and obtain the corresponding adjustment label values. Calculate the percentage of each adjusted label value in N historical collaborative work records; The adjustment judgment value is obtained by multiplying each percentage by its corresponding adjustment label value and summing the results.

9. A freight container transportation dispatching and management system according to claim 8, characterized in that, The process of obtaining the adjustment tag value corresponding to the historical collaborative work record is as follows: The duration of the synergistic chain effect after each historical operation is executed and the frequency of gantry crane rescheduling are standardized and then assembled into an influence vector. The number of clusters K is determined by the elbow rule, and the influence vector is clustered into K clusters based on the K-means clustering algorithm. The mean of the influence vector of each cluster is calculated. Calculate the magnitude of the mean influence vector of each cluster group, sort the cluster groups in ascending order of magnitude, and use the sequence number as the adjustment label value corresponding to each historical collaborative work record in the cluster group.

10. A freight container transportation dispatching and management system according to claim 8, characterized in that, The process of adjusting the collaborative operation instruction if the adjustment judgment value is greater than the set judgment threshold is as follows: The collaborative task adjustment instruction corresponding to the historical collaborative task record with the smallest similarity value is used as the new collaborative task instruction.

Images