Intelligent scheduling method and system for ship lock considering ship classification and dangerous goods isolation

Abstract

The application provides a ship lock intelligent scheduling method and system considering ship classification and dangerous goods isolation. The ship lock intelligent scheduling method considering ship classification and dangerous goods isolation comprises the following steps: S1, obtaining information of a to-be-dispatched ship and information of a lock chamber, and constructing a conventional ship information list ListN and a dangerous ship information list ListD based on a danger level of the to-be-dispatched ship; S2, obtaining a danger level that needs to be separately locked according to a water conservancy hub management method of the lock chamber, modifying a safety size of a ship that needs to be separately locked in the ListD so as to fill the lock chamber; S3, taking the ListN and the ListD as input lists ListI respectively, calculating a global highest scheduling score of the ListI based on a safety size, an arrival time and a type of the ship, and selecting a global optimal scheduling scheme corresponding to a higher global highest scheduling score in the ListN and the ListD as a final scheduling scheme. The application realizes the joint target of improving the utilization rate of the lock chamber, reducing the waiting time of the ship, balancing the priority of the ship and isolating the dangerous goods ship.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of water transportation intelligent management and water conservancy informatization, and particularly relates to a ship lock intelligent scheduling method and system considering ship classification and dangerous goods isolation. BACKGROUND

[0002] Ship lock scheduling is crucial in water transportation, and its rationality is directly related to the operation efficiency and economic benefits of the ship lock. Generally, the ship lock scheduling strategy should follow the general principle of "first-come-first-served, priority to key points, and consideration of others", on the basis of maximizing the utilization of the lock chamber area and minimizing the waiting time of the ships, to maximize the passing capacity per unit time, and economically and reasonably arrange the lock times and schedule the ships. The scheduling strategy needs to consider various factors, including the arrival time, type, size, whether carrying dangerous goods of the ships, and the size of the lock chamber; at the same time, maximizing the utilization rate of the lock chamber also needs to solve the ship entering sequence and parking position, which has certain requirements for spatial operation capacity. It is difficult to achieve the optimal solution by manual scheduling, and it is inevitable to cause many subjective components in the results and unreasonable schemes.

[0003] To improve the efficiency of lock chamber navigation on the basis of ensuring the fairness of scheduling scheme, to exclude the subjective factors of dispatchers and to reduce their work burden, researchers have carried out a series of research on automatic scheduling of ship lock. The paper "Research on Ship Scheduling Optimization Model and Method in Large Water Transport Hub Compound Navigation Mode" simplifies the ship lock scheduling problem into a two-dimensional packing problem, sorts the ship lock-in sequence based on the "first come first served" principle, and then "packs" the lock chamber according to the order. The Bottom-Left-Fill algorithm is used to control the packing process to improve the area utilization rate. Although this method realizes the automatic generation of the ship lock scheduling scheme to some extent, the ship scheduling priority only considers the arrival order, and the fixed lock chamber "packing" order cannot ensure the maximum utilization rate of the lock chamber. The ship lock intelligent scheduling system disclosed in Chinese patent application CN112668846A is similar to the paper "Research on Ship Scheduling Optimization Model and Method in Large Water Transport Hub Compound Navigation Mode", but optimizes the "packing" order: the first ship to be scheduled is selected to "pack" the lock chamber, and the first ship is still "packed" according to the arrival time order, and the last selected lock chamber utilization rate is the optimal solution. Although this method alleviates the impact of sorting on improving the utilization rate of the lock chamber to some extent compared to "Research on Ship Scheduling Optimization Model and Method in Large Water Transport Hub Compound Navigation Mode", it only optimizes the first ship to enter the lock, and still cannot optimize the ship lock-in order to maximize the utilization rate of the lock chamber. In contrast, the ship lock scheduling method based on improved multi-objective genetic algorithm disclosed in Chinese patent application CN107992967A converts the scheduling problem into a multi-objective problem with the goal of reducing ship waiting time and maximizing lock chamber utilization rate, and optimizes the ship scheduling order to maximize the utilization rate of the lock chamber by constructing an initial population of different "packing" orders and solving it based on the improved genetic algorithm. Although this method maximizes the utilization rate of the lock chamber under the premise of using arrival order as the fairness benchmark, it does not consider the priority of different types of ships, such as law enforcement ships, regular commercial ships, and emergency ships. Similarly, the ship lock scheduling method described in the paper "Three Gorges Ship Lock Scheduling Method Considering Lock Chamber Area Utilization Rate" also converts the scheduling problem into a multi-objective problem with the goal of reducing ship waiting time and maximizing lock chamber utilization rate, and uses the ant colony algorithm to solve it. Although the ship size is considered in the process of calculating the ship priority, the ship type is not considered. The ship scheduling method described in the paper "Application of Best fit algorithm in Three Gorges ship lock scheduling" determines the lock-in priority based on the arrival time and type of the ship, and uses the Best fit algorithm to solve the scheduling scheme with the goal of maximizing the utilization rate of the lock chamber. Although this method considers the ship type when determining the scheduling priority, the Best Fit algorithm used is a greedy algorithm, and the solution is not optimized for the lock-in order, which is a local optimal solution.The ship lock scheduling method described in the paper "Research on Ship Lock Scheduling Simulation Optimization Based on Ship Type Standardization" also comprehensively considers the arrival time and type of the ship, but the ship lock-in sequence basically serves to maximize the lock chamber utilization rate, and only when the ship waiting time exceeds a certain threshold, the adjustment into the lock is considered first, that is, the priority of most ships into the lock is irrelevant to the arrival time and type, which is difficult to ensure fairness.

[0004] In summary, the existing ship lock scheduling methods do not consider the type of ship or the solution cannot balance the lock chamber utilization rate and ship priority, resulting in the imbalance between the ship lock navigation efficiency and the scheduling fairness. In addition, there is no method to handle the dangerous ship scheduling problem, which cannot handle special cases in real scenarios. Therefore, it is necessary to develop a ship scheduling method that can automatically optimize the ship scheduling sequence and its position in the lock chamber based on the type of ship, ship size, waiting time, lock chamber size, and dangerous level information, with the joint goals of improving lock chamber utilization rate, reducing waiting time, and balancing ship priority. SUMMARY

[0005] The purpose of the present application is to overcome the defects of the prior art, provide an intelligent ship lock scheduling method and system considering ship classification and dangerous goods isolation, based on the type, size, waiting time, lock chamber size, and dangerous level information of the ship, automatically optimize the ship scheduling sequence and its position in the lock chamber, achieve the joint goals of improving lock chamber utilization rate, reducing ship waiting time, balancing ship priority, and isolating dangerous goods ships.

[0006] To solve the above technical problems, in a first aspect, the present application provides an intelligent ship lock scheduling method considering ship classification and dangerous goods isolation, comprising:

[0007] S1, obtaining the information of the ship to be arranged and the information of the lock chamber, the information of the ship to be arranged including safe size, arrival time, type and dangerous level, the information of the lock chamber including lock chamber size, constructing a conventional ship information list ListN and a dangerous ship information list ListD based on the dangerous level of the ship to be arranged;

[0008] S2, according to the water conservancy hub management method where the lock chamber is located, obtaining the dangerous level that needs to be separately locked, modifying the safe size of the ship that needs to be separately locked in ListD so that it occupies the lock chamber;

[0009] S3, respectively taking ListN and ListD as input list ListI, based on the safe size, arrival time, and type of the ship, calculating the global highest scheduling score of ListI, selecting the global optimal scheduling scheme corresponding to the higher global highest scheduling score in ListN and ListD as the final scheduling scheme.

[0010] In some embodiments, step S3 comprises:

[0011] S31, constructing K copies of ListI SetList = {ListI1, ListI2 …… ListIK};

[0012] For any list ListIk (1≤k≤K), constructing a corresponding counter Mk=0, and randomly shuffling the elements in ListIk;

[0013] Calculating the scheduling score of ListIk, taking the scheduling score as the historical highest score of ListIk, and taking the scheduling scheme as the historical optimal scheduling scheme of ListIk.

[0014] In some embodiments, step S3 comprises:

[0015] S32, for any list ListIk, randomly shuffling the first Rk elements in ListIk, Rk being greater than the number of vessels that can be loaded into the lock chamber in the historical optimal scheduling scheme;

[0016] Calculating the scheduling score of ListIk, if the scheduling score is lower than the previous scheduling score, then ListIk returns to the state before the first Rk elements are shuffled, and the counter Mk is incremented by 1, otherwise, taking the scheduling score as the historical highest score, and taking the scheduling scheme as the historical optimal scheduling scheme;

[0017] S33, repeating the random selection of the list with the highest historical highest score in the K copies of ListI for n times, and executing step S32 on the selected list;

[0018] S34, for any list ListIk in SetList, determining whether the counter Mk of ListIk is greater than a preset number of times α, if yes, resetting the elements and the counter Mk of ListIk, and updating the corresponding historical highest score and the historical optimal scheduling scheme;

[0019] S35, calculating the global highest scheduling score of ListI and the corresponding global optimal scheduling scheme;

[0020] S36, repeating steps S32 to S35 until the number of repetitions reaches a preset upper limit β, or the global highest scheduling score remains unchanged for a number of times reaching a preset upper limit γ.

[0021] In some embodiments, the method of calculating the scheduling score of ListIk comprises: regarding the lock chamber as a two-dimensional container, regarding each vessel as a two-dimensional box, using the Bottom-Left-Fill method, loading the vessels in ListIk into the lock chamber in order, and calculating the sum of the weights of the vessels loaded into the lock chamber as the scheduling score, wherein the weight of a vessel is calculated according to the safety size, arrival time and type of the vessel.

[0022] In some embodiments, the vessel Si WeightS i = L Si ×W Si ×Exp(T0 - T Si )×ScoreT(C Si ), where L Si For ship S i The safe length, W Si For ship S i The safe width, Exp() represents the exponent of a natural number, T0 represents the current time, T Si Indicates ship S i The arrival time is specified, and ScoreT() indicates that the preset score corresponding to the type of ship is obtained.

[0023] In some embodiments, step S33 includes:

[0024] S331. Repeat steps S332 to S333 n times;

[0025] S332. For any list ListIk in SetList, calculate its probability distribution range Range(k) = [ / , / ), where when i≠0, Score(i) represents the current scheduling score of the i-th list, and when i==0, Score(0)=0;

[0026] S333. Randomly generate a number η in the range [0,1), determine the range Range(i) to which η belongs, select the list ListIi corresponding to the range, and perform step S32 on this list.

[0027] In some embodiments, step S33 includes:

[0028] S331. Repeat step S332 n times;

[0029] S332. Extract the list of the top 50% of the highest historical scores from SetList, and randomly select one of the lists, ListIk, and perform step S32 on this list.

[0030] Secondly, the present invention provides a system for implementing the aforementioned intelligent lock scheduling method considering ship classification and dangerous goods isolation, comprising:

[0031] The data acquisition module is used to obtain information about the vessels waiting to be released and the lock chamber.

[0032] A data processing module is configured to construct a normal ship information list ListN and a dangerous ship information list ListD, and modify the safety size of a ship in the ListD that needs to pass through the lock separately so as to fill the lock chamber.

[0033] A score calculation module is configured to calculate the global highest dispatch score of the ListN and the ListD.

[0034] A scheme selection module is configured to compare the global highest dispatch scores of the ListN and the ListD, and select the global optimal dispatch scheme corresponding to the higher global highest dispatch score as the final dispatch scheme.

[0035] In a third aspect, the present application provides a non-transitory computer readable storage medium for storing a computer program or instructions, which, when executed by a computer, cause the ship lock intelligent dispatch method considering ship classification and dangerous goods isolation to be implemented.

[0036] In a fourth aspect, the present application provides a computer program product, which comprises computer instructions; when part or all of the computer instructions are run on a computer, the ship lock intelligent dispatch method considering ship classification and dangerous goods isolation is executed.

[0037] The present application has the following beneficial effects:

[0038] 1. The present application automatically optimizes the ship dispatch order and the position of the ship in the lock chamber based on the type, size, waiting time, lock chamber size and dangerous level information of the ship, realizes the joint goals of improving the utilization rate of the lock chamber, reducing the waiting time of the ship, balancing the priority of the ship and isolating the dangerous goods ship.

[0039] 2. The present application introduces the safety area of the ship as a weight factor to avoid the dispatch scheme being biased towards small ships, ensures fairness as a prerequisite, and improves the space utilization rate of the lock chamber; in combination with the improved Bottom-Left-Fill packing algorithm and the artificial bee colony optimization mechanism, the order and position of the ship are dynamically adjusted to realize a more optimal loading layout, thereby improving the number of ships passing through in a single navigation and reducing the vacancy rate.

[0040] 3. The present application introduces an exponentially increasing waiting time factor in the design of the weight of the ship to ensure that the ship with longer waiting time is dispatched preferentially, avoids the long-term retention of individual ships, and especially solves the problem that the dangerous ship is easily delayed when the number of dangerous ships is small, thereby significantly improving the fairness and overall efficiency of the ship passing through the lock.

[0041] 4. The present application selects the list with the higher historical highest score by probability, resets the list that cannot be further optimized, accelerates the convergence speed of the artificial bee colony algorithm, and avoids local optimum.

[0042] 5. The application lists dangerous goods ships and conventional ships in different lists to solve scheduling schemes respectively, and selects a scheduling scheme with a higher score from them to realize scheduling isolation of dangerous goods ships and conventional ships; in the process of solving the scheduling scheme for the list of new goods ships, the application realizes the separate lockage scheduling of specific dangerous ships by modifying the safety size of the separate lockage dangerous goods ship to the size of the lock chamber. BRIEF DESCRIPTION OF DRAWINGS

[0043] Figure 1 The flow chart of the method of the application is shown in the figure.

[0044] Figure 2 The figure shows a schematic diagram of ship lockage simulation using the bin packing problem.

[0045] Figure 3 The figure shows a schematic diagram of the influence of the bin packing order on the Bottom-Left-Fill result.

[0046] Figure 4 The figure shows a schematic diagram of the influence of the ship area factor on the scheduling score.

[0047] Figure 5 The figure shows a schematic diagram of the application of the application in a certain water control project. DETAILED DESCRIPTION

[0048] In order to make the technical problems, technical solutions and beneficial effects of the present application clearer, the present application will be further described in detail below with reference to the accompanying drawings and examples. It should be understood that the specific examples described herein are only used to explain the present application and do not limit the present application.

[0049] As shown in Figure 1 , the application provides a ship lock intelligent scheduling method considering ship classification and dangerous goods isolation, comprising:

[0050] S1, obtaining information of a ship to be arranged and information of a lock chamber, the information of the ship to be arranged including a safety size, an arrival time, a type and a dangerous level, the information of the lock chamber including a size of the lock chamber, and constructing a conventional ship information list ListN and a dangerous ship information list ListD based on the dangerous level of the ship to be arranged.

[0051] Step S1 includes:

[0052] S11, obtaining a depth L z and a width W z of an inner wall of the lock chamber, and obtaining a current time T0;

[0053] S12, initializing the conventional ship information list ListN and the dangerous ship information list ListD, and for each ship S i waiting to pass through the lock, performing steps S121-S123;

[0054] S121, calculate the safety length L i of the ship S Si = ship length + allowance length, calculate the safety width W i of the ship S Si = ship width + allowance width, wherein the allowance length and the allowance width are determined according to the General Design Specification for Sea Port;

[0055] S122, obtain the arrival time T i , the type C Si and the danger level E Si of the ship S Si , and construct the information element of the ship S i = (safety length L Si , safety width W Si , arrival time T Si , type C Si , danger level E Si );

[0056] S123, if the danger level E Si is “none”, store the information element of the ship S i in the list ListN; otherwise, store the information element of the ship S i in the list ListD.

[0057] For ships carrying dangerous goods, each lock management method usually requires them to pass through the lock separately from regular ships, so in the present application, dangerous ships and regular ships are stored in two different lists and processed separately.

[0058] S2, according to the water conservancy project management method in which the lock chamber is located, obtain the danger level that needs to pass through the lock separately, and modify the safety size of the ship in ListD that needs to pass through the lock separately so that it fills the lock chamber, i.e., modify the safety size of the ship in ListD that needs to pass through the lock separately as: safety length L Si = L z , safety width W Si = W z .

[0059] For ships with a high danger coefficient, the lock management method usually classifies them into a specific danger level, and requires ships of this danger level to pass through the lock separately.

[0060] S3, respectively take ListN and ListD as input list ListI, calculate the global highest scheduling score of ListI based on the safety size, arrival time and type of the ship, and select the global optimal scheduling scheme corresponding to the higher global highest scheduling score of ListN and ListD as the final scheduling scheme.

[0061] Step S3 comprises:

[0062] S31, constructing K copies of ListI SetList = {ListI1, ListI2 …… ListIK}; in this embodiment, 10 ship information lists are copied, i.e. K = 10;

[0063] For any list ListIk (1≤k≤K), the following steps are performed:

[0064] Initialize its corresponding count Mk = 0, historical highest score OptScorek = 0 and historical optimal scheduling scheme Pk, and randomly shuffle the elements in ListIk, i.e. shuffle the order of ships in ListIk;

[0065] Calculate the scheduling score of ListIk, take the scheduling score as the historical highest score of ListIk, i.e. OptScorek = Scorek, and take the scheduling scheme as the historical optimal scheduling scheme Pk of ListIk, the scheduling scheme being the ship loading order and loading position.

[0066] The method for calculating the scheduling score of ListIk comprises: Figure 2 as shown, the lock chamber is regarded as a two-dimensional container with a height L z and a width W z , each ship is regarded as a two-dimensional box with a height L Si and a width W Si , the Bottom-Left-Fill method is adopted, the ships in ListIk are loaded into the lock chamber in order, and the sum of the weights of the ships loaded into the lock chamber is calculated as the scheduling score, wherein the weight of a ship is calculated according to the safety size, arrival time and type of the ship, and the weight WeightS i of ship S i = L Si ×W Si ×Exp(T0 -T Si )×ScoreT(C Si ), wherein L Si is the safety length of ship S i , W Si is the safety width of ship S i , Exp() represents the exponent of a natural number, T0 represents the current time, T Si represents the arrival time of ship S i , and ScoreT() represents obtaining the preset score corresponding to the type of the ship, the higher the lock priority of the ship type, the higher the score.

[0067] The scheduling score sums the ship weights that are loaded into the lock chamber. A high score encourages scheduling ships with long waiting times and high priorities into the lock.

[0068] The ship weight of the present application introduces the safety area of the ship, avoiding the scheduling solution to preferentially schedule small ships into the lock, resulting in large ships being delayed. Figure 4 As shown, without considering the area of the ship, i.e., the ship weight WeightS i = Exp(T0 - T Si )×ScoreT(C Si ), even though the ship waiting time, priority and lock chamber utilization rate in the right scheduling solution are higher than those in the left scheduling solution, the scheduling score of the right side is still lower than that of the left side; and after introducing the area of the ship into the ship weight, the scheduling score of the right side is normally higher than that of the left side.

[0069] The present application uses the Bottom-Left-Fill method to solve the packing problem. Although this method has high computational efficiency and can simulate the process of ships entering the lock, the solution result is affected by the order of ships entering the lock, as shown in Figure 3 With the same four boxes, using the order shown in (a) in Figure 3 based on the Bottom-Left-Fill method, the green box cannot be loaded, while using the order shown in (b) in Figure 3 all boxes can be successfully loaded. Therefore, during the process of iteratively solving the optimal scheduling solution based on the artificial bee colony, the order of ships entering the lock needs to be changed.

[0070] S32, for any list ListIk, the following steps are performed:

[0071] The first Rk elements in ListIk are randomly shuffled, Rk is greater than the number of ships that can be loaded into the lock chamber in the historically optimal scheduling solution, and is less than the number of ships to be arranged; in the present embodiment, Rk = min(the number of ships that can be loaded into the lock chamber in the historically optimal scheduling solution × 1.2, the number of ships to be arranged), wherein min() represents taking the minimum value, and Rk is taken in an upward manner, i.e., if Rk = 4.8, Rk = 5 is taken;

[0072] The scheduling score of ListIk is calculated, if the scheduling score is lower than the previous scheduling score, ListIk is returned to the state before the first Rk elements are shuffled, and the count Mk is increased by 1, otherwise, the scheduling score is taken as the highest score, i.e., OptScorek = Scorek is updated, and the scheduling solution is taken as the historically optimal scheduling solution, i.e., the optimal scheduling solution Pk is updated to the current solution;

[0073] In order to speed up the solution of the optimal scheduling solution, the present application increases the number of "evolutions" for the list with a higher historically highest score:

[0074] S33, randomly select n times from the list with the highest historical score among the K copies of ListI, and execute step S32 on the selected list.

[0075] There are many ways to select the list with the highest historical score among the K copies of ListI, such as the strategy of following bees selecting leader bees in the artificial bee colony algorithm:

[0076] Step S33 includes: S331, repeat steps S332-S333 n times;

[0077] S332, for any list ListIk in SetList, calculate its probability distribution range Range(k)=[ / , / ], where Score(i) represents the current scheduling score of the i-th list when i≠0, and Score(0)=0 when i==0;

[0078] S333, randomly generate a number η in the range [0, 1), and determine the range Range(i) to which η belongs, select the list ListIi corresponding to the range, and execute step S32 on this list.

[0079] This method makes the probability of further evolution of the list with the highest historical score greater.

[0080] For example, step S33 can be simplified as:

[0081] S331, repeat step S332 n times;

[0082] S332, extract the top 50% of lists with the highest historical score from SetList, and randomly select one list ListIk, execute step S32 on this list.

[0083] S34, for any list ListIk in SetList, determine whether the count Mk of ListIk is greater than a preset number α, if yes, reset its elements and count Mk, and update the corresponding historical highest score and historical optimal scheduling scheme; in this embodiment, α=10, i.e. if the ship information list fails to improve the scheduling score for 10 consecutive times, reset its state.

[0084] Step S34 includes:

[0085] S341, for any list ListIk in SetList, if its count Mk is greater than a preset number α, execute steps S342-S344 as follows:

[0086] S342, randomly shuffle the order of elements in ListIk, reset the count Mk=0;

[0087] S343, calculate the dispatch score of ListIk;

[0088] S344, if the current dispatch score Scorek is greater than OptScorek, update OptScorek=Scorek, and update the optimal dispatch scheme Pk as the current scheme.

[0089] S35, count the global highest dispatch score of ListI and the corresponding global optimal dispatch scheme;

[0090] S36, repeat steps S32 to S35 until the number of repetitions reaches a preset upper limit β, or the number of times the global highest dispatch score remains unchanged reaches a preset upper limit γ. In an embodiment, the preset upper limit β=10000, and the preset upper limit γ=20.

[0091] S37, select the global optimal dispatch scheme corresponding to the higher of the global highest dispatch score of ListN and ListD as the final dispatch scheme.

[0092] The application separately lists dangerous ships and conventional ships in different ship information lists for solving, avoiding the mixing of the two in the same lock chamber; by selecting the dangerous ship information list and the conventional ship information list with higher dispatch scores, it is determined whether the current lock is for dangerous ships or conventional ships. The dispatch score encourages ships with longer waiting times to enter the lock first, and the weight of a single ship increases exponentially with the waiting time, thus ensuring the fairness of the dispatch to some extent and avoiding the situation that the number of dangerous ships is small and the lock chamber cannot be "filled" to pass through the lock for a long time.

[0093] It should be noted that an excellent dispatch scheme should, on the basis of considering the priority of different types of ships passing through the lock, reduce the waiting time of the ships as much as possible and improve the utilization rate (navigation rate) of the lock chamber.

[0094] The method quantifies the comprehensive consideration of the dispatch scheme for ship types, waiting time and lock chamber utilization rate by designing an objective function (dispatch score). And by modifying the artificial bee colony algorithm, the dispatch scheme with the highest dispatch score is solved. In the process of solving the dispatch scheme, the ship entering the lock process and the placement position problem are involved, which can be simplified as a two-dimensional packing problem, that is, each ship is regarded as a rectangular cargo and is packed into the "box" of the lock chamber, as shown in Figure 2 .

[0095] The idea of solving the optimal dispatch scheme based on the artificial bee colony is as follows Figure 1S31 is an initialization process, first randomly initializing K different ordered ship information lists, and respectively solving the scheduling ("bin packing") scheme and scheduling score, wherein the scheduling score comprehensively evaluates the type of the entering ship, the waiting time and the lock chamber utilization rate; S32 is a random "evolution" process, which attempts to adjust the order of the K ship information lists in a small range to obtain a higher scheduling score; S33 is similar to the "attention" mechanism of machine learning, for the ship information list with the highest historical score, further accelerating the "evolution" process; S34 resets the state for the list whose score cannot be further increased; the optimal scheduling scheme can be gradually iterated by repeatedly executing steps S32-S35, and S36 is the iteration termination condition, i.e. the global highest scheduling score no longer increases or the iteration reaches a certain number of times.

[0096] As shown in Figure 5 The application has been successfully applied to the ship lock scheduling module of the intelligent operation management system of a certain hydropower station in April 2024. Practice shows that the method can effectively support the intelligent generation of the ship lock scheduling scheme, significantly reduce the work burden of the scheduling personnel, and greatly improve the ship lock navigation efficiency, thereby providing strong support for the intelligent operation of the hydropower station.

[0097] The above-described embodiments are only used to illustrate the technical solutions of the present application, and not to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacements for part of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application, and should be included in the protection scope of the present application.

Claims

1. A smart lock scheduling method considering ship classification and dangerous goods segregation, characterized in that: include: S1. Obtain information on vessels to be released and lock chambers. The information on vessels to be released includes safe dimensions, arrival time, type, and hazard level. The information on lock chambers includes lock chamber dimensions. Construct a list of regular vessel information (ListN) and a list of dangerous vessel information (ListD) based on the hazard level of the vessels to be released. S2. Based on the management regulations of the water conservancy hub where the lock chamber is located, obtain the danger level that requires separate passage through the lock, and modify the safety dimensions of the vessels that require separate passage through the lock in ListD so that they fill the lock chamber. S3. Using ListN and ListD as input lists ListI respectively, calculate the global highest scheduling score of ListI based on the ship's safe size, arrival time, and type. Select the global optimal scheduling scheme corresponding to the higher global highest scheduling score between ListN and ListD as the final scheduling scheme. Step S3 includes: S31. Construct K copies of ListI: SetList = {ListI1, ListI2, ..., ListIK}; For any list ListIk, where 1≤k≤K, construct a corresponding counter Mk=0, and randomly shuffle the elements in ListIk; Calculate the scheduling score of ListIk, take the scheduling score as the highest historical score of ListIk, and take the corresponding scheduling scheme as the best historical scheduling scheme of ListIk. S32. For any list ListIk, randomly shuffle the first Rk elements in ListIk, where Rk is greater than the number of ships that can be loaded into the lock chamber in the historical best scheduling scheme. Calculate the scheduling score of ListIk. If the scheduling score is lower than the previous scheduling score, ListIk reverts to the state before the first Rk elements were shuffled, and the count Mk is incremented by 1. Otherwise, the scheduling score is taken as the highest historical score, and the scheduling scheme is taken as the best historical scheduling scheme. S33. Randomly select the list with the highest historical score from the K copies of ListI n times, and execute step S32 on the selected list; S34. For any list ListIk in SetList, determine whether the count Mk of ListIk is greater than the preset number of times α. If so, reset its elements and count Mk, and update the corresponding historical highest score and historical best scheduling scheme. S35. Calculate the highest global scheduling score of ListI and the corresponding global optimal scheduling scheme; S36. Repeat steps S32 to S35 until the number of repetitions reaches the preset upper limit β, or the number of times the global highest scheduling score remains unchanged reaches the preset upper limit γ.

2. The intelligent lock scheduling method considering ship classification and dangerous goods isolation according to claim 1, characterized in that: The method for calculating the scheduling score of ListIk includes: treating the lock chamber as a two-dimensional container and each ship as a two-dimensional box, using the Bottom-Left-Fill method to load the ships in ListIk into the lock chamber in sequence, and calculating the sum of the weights of the ships loaded into the lock chamber as the scheduling score. The weight of the ship is calculated based on the ship's safe size, arrival time, and type.

3. The intelligent lock scheduling method considering ship classification and dangerous goods isolation according to claim 2, characterized in that: Ship S i WeightS i = L Si ×W Si ×Exp(T0 - T Si )×ScoreT(C Si ), where L Si For ship S i The safe length, W Si For ship S i The safe width, Exp() represents the exponent of a natural number, T0 represents the current time, T Si Indicates ship S i The arrival time is specified, and ScoreT() indicates that the preset score corresponding to the type of ship is obtained.

4. The intelligent lock scheduling method considering ship classification and dangerous goods isolation according to claim 1, characterized in that: Step S33 includes: S331. Repeat steps S332 to S333 n times; S332. For any list ListIk in SetList, calculate its probability distribution range Range(k) = [ / , / ), where when i≠0, Score(i) represents the current scheduling score of the i-th list, and when i=0, Score(0)=0; S333. Randomly generate a number η in the range [0,1), determine the range Range(i) to which η belongs, select the list ListIi corresponding to the range, and perform step S32 on this list.

5. The intelligent lock scheduling method considering ship classification and dangerous goods isolation according to claim 1, characterized in that: Step S33 includes: S331. Repeat step S332 n times; S332. Extract the list of the top 50% of the highest historical scores from SetList, and randomly select one of the lists, ListIk, and perform step S32 on this list.

6. A system for implementing the intelligent lock scheduling method considering ship classification and dangerous goods segregation as described in any one of claims 1 to 5, characterized in that, include: The data acquisition module is used to obtain information about the vessels waiting to be released and the lock chamber. The data processing module is used to construct the regular vessel information list ListN and the dangerous vessel information list ListD, and to modify the safety dimensions of vessels in ListD that need to pass through the lock separately so that they fill the lock chamber. The scoring calculation module is used to calculate the highest global scheduling score for ListN and ListD; The scheme selection module is used to compare the highest global scheduling scores of ListN and ListD, and select the globally optimal scheduling scheme corresponding to the one with the higher highest global scheduling score as the final scheduling scheme.

7. A non-transitory computer-readable storage medium, characterized in that: Used to store computer programs or instructions, which, when executed by a computer, enable the implementation of the intelligent lock scheduling method considering ship classification and dangerous goods isolation as described in any one of claims 1-5.

8. A computer program product, characterized in that, The computer program product includes computer instructions; when some or all of the computer instructions are run on a computer, the intelligent lock scheduling method considering ship classification and dangerous goods isolation as described in any one of claims 1-5 is executed.

Images