Scheduling method and system based on ship traffic flow prediction model

Abstract

The invention discloses a scheduling method and system based on a ship traffic flow prediction model, belongs to the technical field of multi-step double-route ships, and is used for solving the problems that in the operation process of an existing multi-step double-line type ship lock, due to the fact that a ship arrangement plan is unreasonable, the utilization efficiency of lock chambers is low, and the cost is low. And the overall operation efficiency and the collaboration of the passing of the fleet are difficult to improve. The method comprises the following steps: carrying out passing analysis on a ship queue of a wait-to-pass water area under a time factor and a queue length factor, and determining a ship queue state; performing dynamic and static data statistics on all ships in the wait-to-pass water area to obtain dynamic attributes and static attributes of each ship; performing interval division of ship flow on the ship queue in the waiting passing area to obtain an initial ship interval; performing target optimization scheduling processing on the ship flow in the ship interval to obtain a final interval ship flow; and a ship flow scheduling planning scheme based on the multi-step double-linear ship lock is generated.

Description

Technical Field

[0001] The present application relates to the field of multi-level dual-route ships, and in particular to a scheduling method and system based on a ship traffic flow prediction model. Background Art

[0002] Inland waterway shipping has the advantages of large transport volume, low energy consumption and less pollution, and occupies an important position in ship transportation.

[0003] In order to improve the current capacity of inland waterways, according to the current waterway hydrological conditions, ship traffic flow, etc., the types of locks are divided into single-step single-line locks, single-step multi-line locks, multi-step single-line locks and multi-step multi-line locks. Different types of locks correspond to different ship entry and exit processes, so the corresponding lock scheduling rules are also different.

[0004] A multi-step double-line ship lock refers to a ship lock in the waterway where multiple steps are formed due to the different height differences in the inland river basin. Each step ship lock has two lock chambers, which is a multi-step double-line ship lock. The existing multi-step double-line ship lock usually adopts a dam-divided scheduling method, that is, the two locks do not interfere with each other. However, due to the differences in the lock chamber space and operation processes of different locks, the operation time of ships entering and exiting the two step ship locks has a large difference, which makes it easy for ships to be backlogged and congested in the waterway. Summary of the invention

[0005] The embodiments of the present application provide a scheduling method and system based on a ship traffic flow prediction model, which is used to solve the following technical problems: During the operation of the existing multi-step double-line ship lock, due to the unreasonable ship arrangement plan, the utilization efficiency of the lock chamber is low, and it is difficult to improve the overall operation efficiency and coordination of the fleet passage.

[0006] The present application embodiment adopts the following technical solutions:

[0007] On the one hand, an embodiment of the present application provides a scheduling method based on a ship traffic flow prediction model, including: according to a ship scheduling time threshold, performing a traffic analysis on a ship queue in a waiting water area under time factors and team leader factors to determine the state of the ship queue; if the ship queue state is a ship congestion state, performing dynamic and static data statistics on all ships in the waiting water area to obtain dynamic attributes and static attributes of each ship; according to the lock chamber space parameters and based on the dynamic attributes and the static attributes, dividing the ship queue in the waiting area into intervals of ship flow to obtain an initial ship interval; through a preset ship traffic flow prediction model, performing target optimization scheduling processing on the ship flow in the initial ship interval to obtain a final interval ship flow; based on the final interval ship flow, generating a ship flow scheduling planning scheme based on a multi-step double-line ship lock.

[0008] The embodiment of the present application can more effectively predict the passage of ships by analyzing the time factors and the captain factors of the ship queue, thereby reducing the waiting time of ships and improving the efficiency of passage in the waters. Through the statistics of dynamic and static data, the characteristics of each ship can be understood more accurately, which is helpful to adopt targeted scheduling strategies to alleviate congestion. Based on the space parameters of the lock chamber and the attributes of the ship, the interval division of the ship flow is helpful to more reasonably allocate the lock chamber resources and improve the efficiency of the lock chamber. Using the preset ship traffic flow prediction model, the ship flow can be targeted and optimized, thereby improving the accuracy of the scheduling plan. It can better adapt to the passage needs of different ships and reduce the uncertainty in the scheduling process. By optimizing ship scheduling, the energy consumption of ships during waiting and low-speed navigation can be reduced. Reasonable ship scheduling can reduce the risk of collision between ships and improve navigation safety.

[0009] In a feasible implementation manner, according to the ship scheduling time threshold, a transit analysis is performed on the ship queue waiting for passage in the waters under the factors of time and length, and the state of the ship queue is determined, which specifically includes: determining the ship scheduling time threshold between the waiting waters and the first ship lock based on the scale factor of the port to be passed; wherein the ship scheduling time threshold is the longest waiting time interval of the first ship lock; through a preset waterway sensing system and based on the ship scheduling time threshold, a preliminary scan is performed on the ships waiting for passage in the waters to determine the previous ship flow information at the previous time node; based on the current time node, the current ship flow information in the waiting waters is scanned and collected; the previous ship flow information is numerically compared and judged with the current ship flow information to determine the flow comparison value; if the flow comparison value is greater than or equal to the first preset threshold, the ship queue state at the current time node is defined as the ship congestion state; if the flow comparison value is less than the first preset threshold, the ship queue state at the current time node is determined as the ship uncrowded state.

[0010] In a feasible implementation manner, if the ship queue state is a crowded ship state, dynamic and static data statistics are performed on all ships in the waiting passage waters to obtain dynamic and static attributes of each ship, specifically including: if the ship queue state is a crowded ship state, the image acquisition route planning and processing of the drone is performed on the waiting passage area to obtain a drone aerial photography route; wherein the drone aerial photography route is a full ship image acquisition route with the shortest time consumption; through the drone and based on the drone aerial photography route, image acquisition processing is performed on all ships in the waiting passage area to obtain a ship static information image; through a preset search box, all ship pixels contained in the ship static information image are rectangularly framed and marked, and based on the foreground and background layers of the ship static information image, the image acquisition route connected to the ship bottom mapping is marked. The water area and the ship area itself are both pixel segmented to obtain the draft area image and the ship image; the ship image itself and the corresponding draft area image are spliced ​​to obtain the image of the area actually occupied by the ship; the key feature information of the image in the image of the area actually occupied by the ship is matched with the key information features in the pre-stored ship declaration information to determine the static attributes of all ships in the waiting waters; wherein the static attributes include: ship type, ship size, ship dimensions, draft depth, cargo type and special ship mark; dynamic information analysis of all ships in the waiting area for passage is performed on the navigation conditions and waiting time, to determine the dynamic attributes of all ships; the dynamic attributes and static attributes are mounted on each ship corresponding to the image of the area actually occupied by the ship.

[0011] In a feasible implementation manner, dynamic information analysis of the navigation conditions and waiting time of all ships in the waiting area is performed to determine the dynamic attributes of all ships, specifically including: obtaining the current flow information of the ships in the waiting area through the drone; performing navigation capacity analysis on the ship type and the draft in the static attributes of each ship to obtain the navigation capacity of each ship; wherein, the smaller the ship type and the shallower the draft, the stronger the navigation capacity corresponding to the ship; performing numerical comparison and judgment between the current flow information and the navigation capacity of each ship to obtain the navigation capacity of each ship. The actual navigation intensity of the ship; wherein, the greater the ratio of the value of the navigation capacity to the data of the current flow information, the higher the actual navigation intensity of the ship; within the ship scheduling time threshold, the waiting time of each ship in the waiting passage area is counted, and the special ship mark in the static attribute is combined with the waiting time for data processing to determine the waiting time for each ship; wherein the waiting time is the time result information after the special ship mark and the waiting time are weighted; the actual navigation intensity and the waiting time of each ship are combined to obtain the dynamic attribute.

[0012] In a feasible implementation manner, according to the lock chamber space parameters, and based on the dynamic attributes and the static attributes, the ship queue in the waiting passage area is divided into intervals of ship flow to obtain an initial ship interval, which specifically includes: extracting the lock chamber space parameters; wherein the lock chamber space parameters at least include: water level drop parameters, space plane length and width parameters, gate number parameters and gate opening number parameters; based on the lock chamber space parameters and the current flow information of the ship, the ship queue in the waiting passage area is divided into diversions to determine the initial queue diversion line; a number of ships located in the left and right edge areas of the initial queue diversion line are determined as pending ships; the dynamic attributes and static attributes of the pending ships are obtained; the waiting time and actual navigation intensity of the dynamic attributes of each pending ship are scored and processed through preset expert scoring rules to obtain a dynamic score; the static attributes of each pending ship are scored and processed The weight ratio of each attribute item is divided to determine the static weight ratio; if the static attribute includes a special ship mark, the special ship mark will be re-weighted based on the current traffic information in the waiting area to obtain a special weight ratio; the static attribute of each pending ship is scored by the static weight ratio and the special weight ratio to obtain a static score; the static score and the dynamic score are integrated and sorted to obtain the attribute priority of each pending ship; according to the priority order of the attribute priority, the ship size of the pending ship is matched with the space parameter of the lock chamber, and the pending ship after the space parameter matching is determined as a passable ship; based on the position information of the passable ship, the initial queue diversion line is corrected, and the initial ship interval containing all passable ships is generated.

[0013] In a feasible implementation manner, before the ship traffic in the initial ship interval is subjected to target optimization scheduling processing through a preset ship traffic flow prediction model to obtain the final interval ship traffic, the method also includes: obtaining historical initial queue diversion line data, historical attribute priority data, historical passable ship data and historical initial ship interval data; using the historical initial queue diversion line data, the historical attribute priority data and the historical passable ship data as inputs of a neural network model; and using the historical initial ship interval data as outputs of the neural network model; performing model training processing on the neural network model according to the training set and validation set of the inputs and based on the output to obtain a ship interval prediction model; generating a ship traffic algorithm based on the number of ships identified by the drone in the historical initial ship interval and the ship scheduling time threshold; combining the ship traffic algorithm with the ship interval prediction model to obtain the ship traffic flow prediction model.

[0014] In a feasible implementation, the ship traffic in the initial ship interval is targeted for optimization and scheduling through a preset ship traffic flow prediction model to obtain the final interval ship traffic, specifically including: obtaining the initial ship interval at the current time node; inputting the initial queue diversion line data, attribute priority data and passable ship data corresponding to the initial ship interval into the ship traffic flow prediction model to obtain the final ship interval; wherein the final ship interval is the interval after the data of the initial ship interval is secondary corrected; based on the ship traffic flow algorithm of the ship traffic flow prediction model, the flow of the final ship interval is calculated to obtain the final interval ship traffic.

[0015] In a feasible implementation, based on the ship flow in the final interval, a ship flow scheduling planning scheme based on a multi-step double-line ship lock is generated, specifically including: identifying all schedulable ships in the ship flow in the final interval; according to the ship arrangement corresponding to the lock chamber space parameters and based on the actual traffic intensity of the static attributes in the schedulable ships, the scheduling information is sent to the schedulable ships respectively to generate the ship flow scheduling planning scheme.

[0016] In a feasible implementation manner, the dispatch information is sent in the form of text messages, radio broadcasts, and radar signal identification.

[0017] On the other hand, the embodiment of the present application also provides a scheduling system based on a ship traffic flow prediction model, including: a queue collection module, which is used to perform a traffic analysis on the ship queue waiting for passage in the water area under time factors and team leader factors according to the ship scheduling time threshold, and determine the ship queue state; a ship data analysis module, which is used to perform dynamic and static data statistics on all ships in the waiting water area if the ship queue state is a ship congestion state, and obtain the dynamic attributes and static attributes of each ship; according to the lock chamber space parameters, and based on the dynamic attributes and the static attributes, the ship queue in the waiting passage area is divided into intervals of ship flow to obtain an initial ship interval; a ship flow calculation module, which is used to perform target optimization scheduling processing on the ship flow in the initial ship interval through a preset ship traffic flow prediction model to obtain the final interval ship flow; a planning scheme generation module, which is used to generate a ship flow scheduling planning scheme based on a multi-step double-line ship lock based on the final interval ship flow.

[0018] The present application provides a scheduling method and system based on a ship traffic flow prediction model. Compared with the prior art, the embodiments of the present application have the following beneficial technical effects:

[0019] 1. Optimize ship traffic efficiency: By analyzing the time factors and team leader factors of the ship queue, the ship traffic situation can be predicted more effectively, thereby reducing the waiting time of ships and improving the efficiency of water passage.

[0020] 2. Alleviate ship congestion: When the ship queue is in a congested state, the statistics of dynamic and static data can more accurately understand the characteristics of each ship, which is helpful to adopt targeted scheduling strategies and alleviate congestion.

[0021] 3. Improve the utilization rate of the lock chamber: Based on the lock chamber space parameters and ship attributes, the ship flow is divided into intervals, which helps to allocate the lock chamber resources more reasonably and improve the utilization efficiency of the lock chamber.

[0022] 4. Prediction and scheduling accuracy: Using the preset ship traffic flow prediction model, the ship traffic can be targeted and optimized, thereby improving the accuracy of the scheduling plan.

[0023] 5. Generate optimized scheduling plan: The final generated ship flow scheduling planning plan based on multi-step double-line locks can better adapt to the passage needs of different ships and reduce the uncertainty in the scheduling process.

[0024] 6. Reduce energy consumption: By optimizing ship scheduling, the energy consumption of ships during waiting and low-speed sailing can be reduced.

[0025] 7. Improve navigation safety: Reasonable ship scheduling can reduce the risk of collision between ships and improve navigation safety.

[0026] 8. Improve the intelligence level of the management system: The implementation of this method helps to improve the intelligence level of the ship traffic management system and achieve more efficient and intelligent ship dispatching. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work. In the drawings:

[0028] Figure 1 A flow chart of a scheduling method based on a ship traffic flow prediction model provided in an embodiment of the present application;

[0029] Figure 2 A schematic diagram of a multi-step double-line ship lock provided in an embodiment of the present application;

[0030] Figure 3A schematic diagram of the structure of a scheduling device based on a ship traffic flow prediction model provided in an embodiment of the present application. DETAILED DESCRIPTION

[0031] In order to enable those skilled in the art to better understand the technical solutions in this application, the technical solutions in the embodiments of this application will be clearly and completely described below in conjunction with the drawings in the embodiments of this application. Obviously, the described embodiments are only part of the embodiments of this application, not all of them. Based on the embodiments of this specification, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of this application.

[0032] The present application embodiment provides a scheduling method based on a ship traffic flow prediction model, such as Figure 1 As shown, the scheduling method based on the ship traffic flow prediction model specifically includes steps S101-S105:

[0033] It should be noted that there are many ports along the multi-step locks. The presence of ships entering and leaving the ports for loading and unloading operations makes the number of ships in the waterway not fixed, showing a dynamic trend, resulting in deviations in the scheduling plans of the upper and lower associated locks. At the same time, due to the influence of water levels on different locks, there are differences in the ability to guide ships through the locks. The existence of internal and external factors seriously affects the daily scheduling effect of the locks. It is now necessary to scan the ships in the waterway through the perception system (big data and the Internet of Things) along the waterway, and collect, analyze and process the data, detect the number of ships in real time, and analyze the static information of ships (size, draft, etc.) in a timely manner, and formulate the lock scheduling plan in real time and dynamically.

[0034] The ship lock generates a scheduling plan based on the existing channel ships. In the process of selecting ships, many issues need to be considered, such as the long waiting time for ships, the crew complaints caused by the relatively unreasonable and unfair lock sorting, and the local environmental instability caused by the excessive number of ships waiting for the lock. It is now necessary to fully consider the impact of ship selection and ship combination scheduling plans on the waiting time and crew satisfaction, and propose a scheduling decision-making technology with comprehensive target optimization to improve the operation efficiency and service quality of the locks and ensure fairness for ships.

[0035] In the process of ship lock dispatching, affected by seasons, cargo flow, crew habits, dispatching methods, different ship flows in the waterway at different time periods, and different types of ships, the division of unified time periods brings thorny problems to ship lock dispatching. Either the ship waiting for the lock affects the ship's waiting time, or the lock waiting for the ship causes a waste of idle resources. Now we need a time-sharing dispatching technology based on time period flow division, which can determine the dispatching time of the ship in different time periods according to different traffic flows and different types of ships, and seek a balance between the waiting time for the ship and the waiting time for the lock.

[0036] S101. According to the ship dispatch time threshold, a passage analysis is performed on the ship queue waiting for passage in the water area under the factors of time and length, and the ship queue status is determined.

[0037] Specifically, it is necessary to first determine the ship dispatching time threshold between the waiting waters and the first ship lock based on the scale factor of the port to be passed. The ship dispatching time threshold is the longest waiting time interval of the first ship lock. Then, through the preset sensing system along the waterway, and based on the ship dispatching time threshold, a preliminary scan of the ships waiting to pass the waters is performed to determine the previous ship flow information at the previous time node.

[0038] Furthermore, based on the current time node, the current ship flow information in the waiting waters is scanned and collected.

[0039] Furthermore, a numerical comparison is performed between the previous ship flow information and the current ship flow information to determine a flow comparison value.

[0040] Further, if the flow comparison value is greater than or equal to the first preset threshold, the ship queue state at the current time node is defined as a ship congestion state. If the flow comparison value is less than the first preset threshold, the ship queue state at the current time node is determined as a ship uncrowded state.

[0041] In one embodiment, Figure 2 A schematic diagram of a multi-step double-line ship lock provided in an embodiment of the present application is shown in FIG. Figure 2 As shown in the figure, it is necessary to use the channel design of the drone to collect the situation of ships preparing to pass through the No. 1 ship lock to the No. 1 lock chamber (the first ship lock) waiting area, and based on the ship scheduling time threshold, to keep track of the waiting situation of the ships at all times, so as to determine the state of the ship queue. Then, after comparing the preliminary estimated previous ship flow information with the current ship flow information, the ship queue state is finally obtained.

[0042] S102: If the ship queue state is a ship congestion state, dynamic and static data statistics are performed on all ships in the waiting waters to obtain dynamic attributes and static attributes of each ship.

[0043] Specifically, if the ship queue state is a crowded state, the image acquisition route planning process of the drone is performed on the waiting area to obtain the drone aerial photography route. The drone aerial photography route is the full ship image acquisition route with the shortest time consumption.

[0044] Furthermore, by using a drone and based on the drone aerial photography route, images of all ships in the waiting area are collected and processed to obtain static information images of the ships.

[0045] Furthermore, it is necessary to use a preset search box to rectangularly frame and mark all the ship pixels contained in the ship static information image, and based on the foreground and background layers of the ship static information image, the draft area connected to the ship bottom mapping and the ship area itself are pixel segmented to obtain the draft area image and the ship itself image.

[0046] Furthermore, the image of the ship itself is spliced ​​with the corresponding draft area image to obtain the image of the area actually occupied by the ship. The key feature information in the image of the area actually occupied by the ship is then matched with the key information features in the pre-stored ship declaration information to determine the static attributes of all ships in the waters waiting for passage. The static attributes include: ship type, ship size, ship dimensions, draft depth, cargo type and special ship marks.

[0047] Furthermore, it is also necessary to analyze the dynamic information of all ships in the waiting area regarding the navigation conditions and waiting time, and determine the dynamic properties of all ships.

[0048] As a feasible implementation method, the current flow information of ships in the waiting passage area can be obtained by drones. Then, the navigation capacity of each ship is analyzed for the ship type and draft depth in the static attributes of each ship to obtain the navigation capacity of each ship. Among them, the smaller the ship type and the shallower the draft, the stronger the corresponding navigation capacity. Then, the current flow information is compared with the navigation capacity of each ship to obtain the actual navigation intensity of each ship. Among them, the greater the ratio of the value of the navigation capacity to the data of the current flow information, the higher the actual navigation intensity of the ship. Within the ship scheduling time threshold, the waiting time of each ship in the waiting passage area is counted, and the special ship mark in the static attribute is combined with the waiting time to determine the waiting time of each ship. Among them, the waiting time is the time result information after the special ship mark and the waiting time are weighted. Finally, the actual navigation intensity and waiting time of each ship are combined to obtain the dynamic attribute.

[0049] Finally, both dynamic attributes and static attributes are mounted on each ship corresponding to the image of the area occupied by the actual ship.

[0050] In one embodiment, the priority of ship scheduling is the key to ship sorting. In the priority rule setting, in the ship queuing rules, the ship is arranged to pass through the lock according to the principle of first come first served and priority priority, and the ship's number and draft are followed. At the same time, the static and dynamic attributes of the waiting ship are considered. Static attributes mainly refer to the intrinsic attributes of the ship, that is, the attributes that will no longer change after the ship declares to pass through the lock, such as ship size, cargo type, etc. The dynamic attributes are attributes that change with time and external environment, such as navigation conditions (when the navigation capacity is sufficient, the weight of the local company's lock tugboat is greater than that of a general cargo ship. When the navigation capacity is not sufficient, the weight of a general cargo ship should be greater than that of a lock tugboat) and waiting time (the longer the waiting time, the greater the weight, and the higher-level ships increase their weight faster as the waiting time increases) By weighing the static and dynamic attributes of the ship, a relatively reasonable and fair ship priority sorting is finally obtained.

[0051] S103, dividing the ship queue in the waiting area into intervals of ship flow according to the lock chamber space parameters and based on the dynamic attributes and the static attributes to obtain an initial ship interval.

[0052] Specifically, the ship gate management system is first used to extract the lock chamber space parameters. The lock chamber space parameters include at least: water level drop parameters, space plane length and width parameters, gate number parameters, and gate opening number parameters. Based on the lock chamber space parameters and the current flow information of the ship, the waiting area is divided into ship queues to determine the initial queue diversion line.

[0053] As a feasible implementation method, intelligent ship lock dispatching rules are constructed based on the rigid conditions such as lock chamber size, ship transit time and heading. Under the premise of ensuring the load balance of multiple lock chambers and reasonable ship delays, the dispatching operation plan for ships waiting to be locked is arranged to maximize the operation efficiency and service level. The level of lock chamber arrangement, ship sequencing and lock chamber gearing is improved.

[0054] Furthermore, several ships located in the left and right edge areas of the initial queue diversion line are determined as pending ships; and the dynamic attributes and static attributes of the pending ships are obtained.

[0055] Furthermore, the waiting time and actual navigation intensity of the dynamic attributes of each pending ship are scored through the preset expert scoring rules to obtain a dynamic score.

[0056] Furthermore, it is necessary to divide the weight ratio of each attribute item of the static attribute of each pending ship to determine the static weight ratio. If the static attribute includes a special ship mark, the special ship mark will be re-weighted based on the current traffic information in the waiting area to obtain a special weight ratio. Finally, the static attribute of each pending ship is scored by the static weight ratio and the special weight ratio to obtain a static score.

[0057] In one embodiment, some ships that are capable of transporting special cargoes or carrying special missions also need to be given priority, so the static attributes of the ship also need to re-weight the special ship marks separately, and then the total static attributes include the static weight ratios of each item and the static score under the special weight ratio.

[0058] Furthermore, the static scores and dynamic scores are integrated and sorted to obtain the attribute priority of each pending ship. Then, according to the priority order of the attribute priority, the ship size of the pending ship is matched with the space parameters of the lock chamber, and the pending ship that passes the space parameter matching is determined as a passable ship. Finally, based on the position information of the passable ship, the initial queue diversion line is corrected and the initial ship interval containing all passable ships is generated.

[0059] In one embodiment, the ship size and space parameters of the lock chamber are matched first for ships with high scores, that is, ships with higher attribute priorities. If they meet the requirements, these ships are used as the last pending ships that can be divided into the left and right edge areas of the initial queue diversion line, thereby ensuring the fairness of ships and the utilization efficiency of the lock chamber to the greatest extent, achieving the maximum balance as much as possible, and only after the final passable ship is determined can the initial ship interval containing all passable ships be generated, that is, the diversion line and diversion area are finally divided.

[0060] S104. Using a preset ship traffic flow prediction model, target optimization scheduling is performed on the ship flow in the initial ship interval to obtain the final interval ship flow.

[0061] Specifically, firstly, the historical initial queue diversion line data, the historical attribute priority data, the historical passable ship data and the historical initial ship interval data are obtained.

[0062] Furthermore, the historical initial queue diversion line data, historical attribute priority data, and historical passable ship data are used as the input of the neural network model. The historical initial ship interval data is used as the output of the neural network model. Then, according to the training set and validation set of the input, and based on the output, the neural network model is trained and processed to obtain the ship interval prediction model.

[0063] Furthermore, a ship flow algorithm is generated based on the number of ships identified by the UAV in the historical initial ship interval and the ship scheduling time threshold.

[0064] Furthermore, it is necessary to combine the ship traffic flow algorithm with the ship interval prediction model to obtain a ship traffic flow prediction model.

[0065] Furthermore, the initial ship interval at the current time node is obtained. The initial queue diversion line data, attribute priority data and passable ship data corresponding to the initial ship interval are input into the ship traffic flow prediction model to obtain the final ship interval. The final ship interval is the interval after the data of the initial ship interval is corrected twice.

[0066] Furthermore, based on the ship traffic flow algorithm of the ship traffic flow prediction model, the flow of the final ship interval is calculated to obtain the final interval ship flow.

[0067] S105. Based on the final interval ship flow, generate a ship flow dispatch planning scheme based on a multi-step double-line ship lock.

[0068] Specifically, all dispatchable ships in the final interval ship flow are first identified. Then, according to the ship arrangement corresponding to the lock chamber space parameters and the actual traffic intensity of the static attributes of the dispatchable ships, the dispatch information is sent to the dispatchable ships respectively to generate a ship flow dispatch planning scheme. The dispatch information is sent in the following ways: SMS, radio broadcast and radar signal identification.

[0069] In addition, the present application also provides a scheduling system based on a ship traffic flow prediction model, such as Figure 3 As shown, the scheduling system 300 based on the ship traffic flow prediction model specifically includes:

[0070] The queue collection module 310 is used to perform a passage analysis of the ship queue waiting for passage in the water area under the factors of time and length according to the ship dispatching time threshold, and determine the ship queue status;

[0071] The ship data analysis module 320 is used to collect dynamic and static data statistics of all ships in the waiting passage area if the ship queue state is a crowded state, and obtain the dynamic attributes and static attributes of each ship; divide the ship queue in the waiting passage area into intervals of ship flow according to the lock chamber space parameters and based on the dynamic attributes and static attributes to obtain the initial ship interval;

[0072] The ship flow calculation module 330 is used to perform target optimization scheduling processing on the ship flow in the initial ship interval through a preset ship traffic flow prediction model to obtain the final interval ship flow;

[0073] The planning scheme generating module 340 is used to generate a ship flow dispatching planning scheme based on a multi-step double-line ship lock based on the final ship flow.

[0074] The embodiment of the present application can more effectively predict the passage of ships by analyzing the time factors and the captain factors of the ship queue, thereby reducing the waiting time of ships and improving the efficiency of passage in the waters. Through the statistics of dynamic and static data, the characteristics of each ship can be understood more accurately, which is helpful to adopt targeted scheduling strategies to alleviate congestion. Based on the space parameters of the lock chamber and the attributes of the ship, the interval division of the ship flow is helpful to more reasonably allocate the lock chamber resources and improve the efficiency of the lock chamber. Using the preset ship traffic flow prediction model, the ship flow can be targeted and optimized, thereby improving the accuracy of the scheduling plan. It can better adapt to the passage needs of different ships and reduce the uncertainty in the scheduling process. By optimizing ship scheduling, the energy consumption of ships during waiting and low-speed navigation can be reduced. Reasonable ship scheduling can reduce the risk of collision between ships and improve navigation safety.

[0075] Each embodiment in this application is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the system storage medium embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0076] The above describes specific embodiments of the present application. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recorded in the claims can be performed in an order different from that in the embodiments and still achieve the desired results. In addition, the processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0077] The above is only an embodiment of the present application and is not intended to limit the present application. For those skilled in the art, the embodiments of the present application may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the embodiments of the present application should be included in the scope of the claims of the present application.

Claims

1. A scheduling method based on a ship traffic flow prediction model, characterized in that: The method comprises: According to the ship dispatch time threshold, the ship queue waiting to pass through the waters is analyzed under the factors of time and length, and the ship queue status is determined; If the ship queue state is a ship congestion state, dynamic and static data statistics are performed on all ships in the waiting waters to obtain dynamic attributes and static attributes of each ship; According to the lock chamber space parameters and based on the dynamic attributes and the static attributes, the ship queue in the waiting passage area is divided into intervals of ship flow to obtain an initial ship interval; By using a preset ship traffic flow prediction model, the ship flow in the initial ship interval is subjected to target optimization scheduling processing to obtain the final interval ship flow; Based on the final interval ship flow, a ship flow scheduling planning scheme based on a multi-step double-line ship lock is generated.

2. A scheduling method based on a ship traffic flow prediction model according to claim 1, characterized in that: According to the ship dispatch time threshold, the ship queue waiting to pass through the waters is analyzed under the factors of time and length, and the ship queue status is determined, including: Based on the scale factor of the port to be passed, determine the ship dispatching time threshold between the waiting passage waters and the first ship lock; wherein the ship dispatching time threshold is the longest waiting time interval of the first ship lock; Through a preset sensing system along the waterway and based on the ship scheduling time threshold, a preliminary scan is performed on the ships waiting to pass through the waters to determine the previous ship flow information at the previous time node; Based on the current time node, scan and collect the current ship flow information in the waiting waters; Compare the previous ship flow information with the current ship flow information to determine a flow comparison value; If the flow comparison value is greater than or equal to a first preset threshold, the ship queue state at the current time node is defined as the ship congestion state; If the flow comparison value is less than the first preset threshold, the ship queue state at the current time node is determined as a ship uncrowded state.

3. The scheduling method based on the ship traffic flow prediction model according to claim 1 is characterized in that: If the ship queue state is a ship congestion state, dynamic and static data statistics are performed on all ships in the waiting waters to obtain dynamic and static attributes of each ship, including: If the ship queue state is a crowded state, the image acquisition route planning process of the drone is performed on the waiting area to obtain a drone aerial photography route; wherein the drone aerial photography route is a full ship image acquisition route with the shortest time consumption; By using the drone and based on the drone's aerial photography route, image acquisition and processing are performed on all ships in the waiting passage area to obtain a static information image of the ships; Through a preset search box, all the ship pixels contained in the ship static information image are rectangularly framed and marked, and based on the foreground and background layers of the ship static information image, the draft area connected to the ship bottom mapping and the ship area itself are pixel segmented to obtain the draft area image and the ship image itself; Performing image stitching processing on the image of the ship itself and the corresponding image of the draft area to obtain an image of the area actually occupied by the ship; Matching the key feature information of the image in the image of the area occupied by the actual ship with the key information features in the pre-stored ship declaration information to determine the static attributes of all ships in the waiting waters; wherein the static attributes include: ship type, ship size, ship dimensions, draft, cargo type and special ship mark; Analyzing the navigation conditions and dynamic information of all ships in the waiting area during the waiting time to determine the dynamic attributes of all ships; The dynamic attributes and the static attributes are mounted on each ship corresponding to the image of the area occupied by the actual ship.

4. The scheduling method based on the ship traffic flow prediction model according to claim 3 is characterized in that: Analyze the navigation conditions and dynamic information of all ships in the waiting area during the waiting time to determine the dynamic attributes of all ships, including: Obtaining, by means of the drone, current flow information of the ship in the waiting area for passage; Performing a navigation capacity analysis on the ship type and the draft in each ship's static attributes to obtain the navigation capacity of each ship; wherein the smaller the ship type and the shallower the draft, the stronger the navigation capacity corresponding to the ship; The current flow information is numerically compared with the navigation capacity of each of the ships to obtain the actual navigation intensity of each of the ships; wherein, the greater the ratio of the navigation capacity to the data of the current flow information, the higher the actual navigation intensity of the ship; Within the ship dispatch time threshold, the waiting time of each ship in the waiting passage area is counted, and the special ship mark in the static attribute is combined with the waiting time to determine the waiting time of each ship; wherein the waiting time is the time result information after the special ship mark and the waiting time are weighted; The actual navigation intensity and the waiting time of each ship are combined to obtain the dynamic attribute.

5. The scheduling method based on the ship traffic flow prediction model according to claim 1 is characterized in that: According to the lock chamber space parameters and based on the dynamic attributes and the static attributes, the ship queue in the waiting passage area is divided into intervals of ship flow to obtain an initial ship interval, specifically including: Extracting the lock chamber space parameters; wherein the lock chamber space parameters at least include: water level drop parameters, space plane length and width parameters, gate quantity parameters and gate opening quantity parameters; Based on the lock chamber space parameters and the current flow information of the ship, the waiting area is divided into ship queues to determine the initial queue diversion line; Determine a number of ships located at the left and right edge areas of the initial queue diversion line as pending ships; Obtaining dynamic attributes and static attributes of the pending ship; Through the preset expert scoring rules, the waiting time and actual navigation intensity of the dynamic attributes of each pending ship are scored to obtain a dynamic score; Divide the attribute items of the static attribute of each pending ship into weight proportions to determine the weight proportions of each static attribute; if the static attribute includes a special ship mark, then re-weight the special ship mark based on the current traffic information in the waiting passage area to obtain a special weight proportion; Scoring the static attributes of each of the pending ships according to the static weight ratios and the special weight ratios to obtain a static score; The static scores and the dynamic scores are integrated and sorted to obtain the attribute priority of each of the pending ships; According to the priority order of the attribute priorities, the ship size of the to-be-determined ship is matched with the space parameters of the lock chamber, and the to-be-determined ship after the space parameter matching passes is determined as a passable ship; Based on the position information of the passable ships, the initial queue diversion line is corrected, and the initial ship section including all passable ships is generated.

6. The scheduling method based on the ship traffic flow prediction model according to claim 1 is characterized in that: Before performing target optimization scheduling processing on the ship traffic flow in the initial ship interval by using a preset ship traffic flow prediction model to obtain the final interval ship traffic flow, the method further includes: Obtain historical initial queue diversion line data, historical attribute priority data, historical passable ship data, and historical initial ship interval data; The historical initial queue diversion line data, the historical attribute priority data, and the historical passable ship data are used as inputs of the neural network model; and the historical initial ship interval data are used as outputs of the neural network model; According to the training set and the validation set of the input amount, and based on the output amount, the neural network model is subjected to model training processing to obtain a ship interval prediction model; Generate a ship flow algorithm based on the number of ships identified by the drone in the historical initial ship interval and the ship scheduling time threshold; The ship traffic flow prediction model is obtained by combining the ship traffic flow algorithm with the ship interval prediction model.

7. The scheduling method based on the ship traffic flow prediction model according to claim 1 is characterized in that: By using a preset ship traffic flow prediction model, the ship traffic in the initial ship interval is subjected to target optimization scheduling processing to obtain the final interval ship traffic, specifically including: Obtaining the initial ship interval at the current time node; Inputting the initial queue diversion line data, attribute priority data and passable ship data corresponding to the initial ship interval into the ship traffic flow prediction model to obtain a final ship interval; wherein the final ship interval is the interval after the data of the initial ship interval is secondary corrected; Based on the ship flow algorithm of the ship traffic flow prediction model, the flow is calculated for the final ship interval to obtain the ship flow in the final interval.

8. The scheduling method based on the ship traffic flow prediction model according to claim 1 is characterized in that: Based on the final interval ship flow, a ship flow dispatching planning scheme based on a multi-step double-line ship lock is generated, specifically including: Identifying all dispatchable vessels in the final interval vessel flow; According to the ship arrangement corresponding to the lock chamber space parameters and based on the actual traffic intensity of the static attributes of the dispatchable ships, the dispatch information is sent to the dispatchable ships respectively to generate the ship flow dispatch planning scheme.

9. The scheduling method based on the ship traffic flow prediction model according to claim 8 is characterized in that: The dispatch information may be sent in the form of text messages, radio broadcasts, and radar signal identification.

10. A dispatching system based on a ship traffic flow prediction model, characterized in that: The system comprises: The queue collection module is used to analyze the passage of the ship queue waiting for passage in the waters under the factors of time and length according to the ship dispatch time threshold, and determine the state of the ship queue; A ship data analysis module is used for, if the ship queue state is a ship congestion state, performing dynamic and static data statistics on all ships in the waiting passage area to obtain dynamic attributes and static attributes of each ship; dividing the ship queue in the waiting passage area into intervals of ship flow according to the lock chamber space parameters and based on the dynamic attributes and the static attributes to obtain an initial ship interval; A ship flow calculation module is used to perform target optimization scheduling processing on the ship flow in the initial ship interval through a preset ship traffic flow prediction model to obtain the final interval ship flow; A planning scheme generating module is used to generate a ship flow dispatching planning scheme based on a multi-step double-line ship lock based on the final interval ship flow.

Images