A method for inventory shipyard vehicle routing optimization based on adaptive large neighborhood search

Abstract

The present application relates to the shipyard yard scheduling technical field, specifically disclose a kind of based on adaptive large neighborhood search and inventory shipyard vehicle route optimization method, comprising the following steps: based on initial inventory and demand to establish mathematical model, in combination with the capacity of vehicle and the distance between warehouse, generate a feasible initial distribution route;Through adaptive large neighborhood search algorithm, the initial distribution route is iteratively optimized to obtain optimal distribution route;Based on maximum inventory constraint, safety inventory constraint and vehicle capacity constraint, optimal distribution route is applied to interference recovery model to split and combine, after multiple iterations, obtain optimal solution under interference state;According to the type and influence of interference, take corresponding recovery strategy to process and determine the scheme of recovery scheduling;Strategy includes vehicle self-help and dispatch new vehicle, when facing interference event, can quickly recover and optimize distribution route, ensure the stability of supply chain and the continuity of production activity.

Description

Technical Field

[0001] This invention relates to the field of shipyard yard scheduling technology, and more specifically, to a method for optimizing the route of inventory shipyard vehicles based on adaptive large neighborhood search and inventory shipyard. Background Technology

[0002] The scope of production logistics activities in manufacturing enterprises includes the entire material flow and management activities from raw material intake to finished product output. Particular emphasis is placed on the importance of production and distribution in supply chain management for fast-moving consumer goods (FMCG) manufacturers. In traditional operating models, enterprise logistics systems include independent basic activities such as production, inventory, and routing, as well as inbound and in-plant logistics. For companies producing large-volume goods, due to high production speeds and large material demands, small warehouses need to be established near the production site and replenished promptly to avoid production stoppages. The inventory levels in these small warehouses are related to the rate of material consumption, constituting an inventory path production problem.

[0003] The shortcomings of existing technology are:

[0004] 1. There is a lack of a general algorithm applicable to all inventory routing production problems. For small and medium-sized cases, since inventory routing production is an NP-hard problem, the optimal or near-optimal solution can be found by simulating natural phenomena or optimization processes. However, the solution process of the algorithm needs to be set according to the problem itself, which is also an area that needs improvement.

[0005] 2. Inventory replenishment strategies are limited by users' inventory levels. Demand is dynamic and random, so reasonable replenishment strategies need to be developed to adapt to random and dynamic demand.

[0006] In view of this, the present invention provides a method for optimizing vehicle routes in a shipyard based on adaptive large neighborhood search and inventory shipyard, in order to solve the above problems. Summary of the Invention

[0007] To overcome the problems in the existing technology, this invention proposes an adaptive large neighborhood search and inventory shipyard vehicle route optimization method to solve the scheduling problem of production logistics vehicles in large manufacturing industries after being disturbed.

[0008] This invention provides a method for optimizing vehicle routes in a shipyard based on adaptive large neighborhood search and inventory, comprising the following steps:

[0009] A mathematical model is built based on initial inventory and demand, and a feasible initial delivery route is generated by combining vehicle capacity and distance between warehouses.

[0010] The initial delivery route is iteratively optimized using an adaptive large neighborhood search algorithm to obtain the optimal delivery route.

[0011] Based on the maximum inventory constraint, safety stock constraint, and vehicle capacity constraint, the optimal delivery route is applied to the disturbance recovery model for decomposition and combination. After multiple iterations, the optimal solution under the disturbance state is obtained.

[0012] Based on the type and impact of the interference, corresponding recovery strategies are adopted to determine the recovery and dispatch plan; strategies include vehicle self-rescue and dispatching additional vehicles.

[0013] As a preferred technical solution of the present invention, a mathematical model is established within a period based on delivery route costs and vehicle costs; the mathematical model includes:

[0014] The objective function is to minimize the total cost, which includes delivery route costs and vehicle costs.

[0015] The constraints are vehicle capacity, warehouse demand, number of vehicles, and routes, among which:

[0016] Vehicle capacity constraint: Each vehicle's delivery volume cannot exceed its maximum capacity.

[0017] Warehouse demand constraint: The demand for each warehouse must be met.

[0018] Vehicle number constraint: The number of vehicles used cannot exceed the total number of available vehicles.

[0019] Route constraint: Each route must start from the main warehouse and return to the main warehouse.

[0020] As a preferred embodiment of the present invention, the logic for obtaining the initial delivery route is as follows:

[0021] Based on a hybrid collaboration framework, warehouse demand is determined by analyzing the initial inventory suitable for warehouse delivery.

[0022] An algorithm for encoding and generating initial solutions based on inventory demand and distance is proposed to obtain feasible solutions.

[0023] The optimal vehicle route is obtained through multiple iterations of adaptive large neighborhood search.

[0024] As a preferred embodiment of the present invention, the mathematical model of the interference event includes:

[0025] When a disruptive event occurs, the objective function is to minimize the total cost, which includes delivery route costs, vehicle costs, and delay costs.

[0026] The constraints are: interference events, vehicle self-rescue, dispatching new vehicles, and safety stock.

[0027] Interference event constraints: Consider the impact of interference events on delivery routes.

[0028] Vehicle self-rescue constraint: When a vehicle encounters interference, it can self-rescue by changing its driving route.

[0029] Additional vehicle deployment constraint: When necessary, additional vehicles can be deployed to serve warehouses that have not been dispatched.

[0030] Safety stock constraint: Ensure that the inventory level in each warehouse is not lower than the safety stock value.

[0031] As a preferred technical solution of the present invention, the warehouse replenishment quantity and vehicle capacity are determined according to the performance-optimized strategy to determine the warehouse distribution combination;

[0032] The delivery combination of warehouses is adjusted based on their distance, and the order of warehouses is determined based on the initial inventory of the warehouses to integrate vehicle delivery routes;

[0033] Determine if any interference incidents have occurred with the delivery vehicle;

[0034] If no disruptions occur to the delivery vehicles, the delivery route will be replanned to minimize the total cost.

[0035] If a disruption event occurs during delivery, the type of disruption event is determined, and the location of each delivery vehicle at the time of the disruption is identified. This information is then used to determine the appropriate recovery strategy and scheduling plan.

[0036] As a preferred technical solution of the present invention, an optimal warehouse early warning value is determined, the optimal warehouse early warning value is marked as a preset early warning value, and different recovery strategies are determined based on the preset early warning value. The recovery strategies include a first strategy and a second strategy.

[0037] If the warehouse warning value is less than or equal to the preset warning value, the first strategy is selected, which means that no new vehicles are dispatched. Instead, the vehicles can be self-rescued by changing the delivery route, such as by dealing with interference caused by factors such as road congestion. This strategy is suitable for interference events with minor impact.

[0038] If the warehouse warning value is greater than the preset warning value, the second strategy is selected, which is to dispatch new vehicles. This is suitable for situations where the waiting time is too long or there is no reasonable new route.

[0039] As a preferred technical solution of the present invention, the hybrid collaborative framework based on large neighborhood search includes: a route generation algorithm, an algorithm for determining warehouse replenishment materials based on performance value adjustment strategy, and a large neighborhood removal and repair operator based on this problem.

[0040] The specific advantages of this invention are as follows:

[0041] This invention proposes a novel stochastic inventory path optimization problem, comprehensively considering material consumption, inventory levels, and vehicle route optimization. It introduces safety stock and warehouse capacity constraints, as well as detailed classification and response strategies for disruptive events. A hybrid collaborative framework based on large neighborhood search is designed, which effectively reduces delivery costs and improves scheduling flexibility and efficiency. In particular, it can quickly recover and optimize delivery routes when facing disruptive events, ensuring the stability of the supply chain and the continuity of production activities. Attached Figure Description

[0042] Figure 1 This application provides an illustration of an internal enterprise production and vehicle route optimization method based on adaptive large neighborhood search and inventory constraints.

[0043] Figure 2 This application provides a disturbance management vehicle self-rescue strategy diagram based on an adaptive large neighborhood search and inventory constraint-based enterprise internal production and vehicle route optimization method;

[0044] Figure 3 This application provides a disturbance management strategy diagram for dispatching new vehicles based on an adaptive large neighborhood search and inventory constraint-based enterprise internal production and vehicle route optimization method.

[0045] Figure 4 This application provides an initial route map for an enterprise internal production and vehicle route optimization method based on adaptive large neighborhood search and inventory constraints.

[0046] Figure 5 This application provides a replanning graph for an enterprise-internal production and vehicle route optimization method based on adaptive large neighborhood search and inventory constraints. Detailed Implementation

[0047] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0048] Figure 1This document illustrates an internal production and vehicle route optimization method for enterprises based on adaptive large neighborhood search and inventory constraints. A reasonable inventory warning, i.e., initial inventory, is established. Based on the initial inventory, the demand for each sub-warehouse can be determined. Combining vehicle capacity and the distance between warehouses, an initial feasible solution, i.e., an initial delivery route, is generated. Then, based on the adaptive large neighborhood search algorithm, the feasible solution is iteratively processed to obtain the optimal solution, and the route corresponding to the optimal solution is designated as the optimal delivery route. To obtain recovery plans in case of route interruptions, scenarios of potential interference during vehicle transportation at the shipyard are analyzed, and the delivery combinations of warehouses are determined by combining and classifying these interferences. Then, different recovery strategies are designed according to different types of interference. In strategies requiring replanning of routes, such as the first and second strategies, the routes are split and combined. The split routes are divided into a set of warehouses with completed deliveries and a set of warehouses with incomplete deliveries. The adaptive large neighborhood search iterative process is then initiated. After each iteration, the incomplete delivery warehouses are recombined into new combinations. Before the set of incomplete delivery warehouses enters the adaptive large neighborhood search iterative process, the memories of completed delivery warehouses, including costs and times, are added. Through multiple iterations, the optimal disturbance recovery path was finally obtained. Each iteration in both parts incorporated the redistribution of warehouse resources to fully utilize vehicle capacity.

[0049] Figure 2 This section explains the vehicle self-rescue strategy in the event of interference with management, and the specific methods are as follows:

[0050] When a transport vehicle encounters a disruption event and does not receive, or does not need, assistance from a replacement vehicle, the disruption can be addressed by altering the vehicle's original route. Strategy 1 is suitable for disruptions with minimal impact on transport, such as vehicle breakdowns requiring waiting, where timely route changes can lead to greater overall delays. Strategy 1 is applied to disruptions in Category 2. The location of the disruption is assumed to be temporary; already delivered locations are excluded, and the delivery route is replanned to minimize costs and warehousing efficiency.

[0051] Figure 3 The specific method for adding new vehicles to the management system to mitigate interference is as follows:

[0052] When a disruption event occurs, if the vehicle's waiting time is too long or there is no reasonable new route to the un-dispatched warehouse (e.g., the route is too long or too costly), a new vehicle is needed to serve the un-dispatched warehouse. The second strategy is applied to disruption events in category 1. Since the total warehouse inventory is limited, it is necessary to pick up the materials from the vehicle experiencing the disruption event. The location of the malfunction is designated as a temporary location, and the new vehicle uses this temporary location as its starting point to replan its route.

[0053] Example 1: A Chinese shipyard uses welding materials for building large cruise ships, including welding rods, welding wires, carbon rods, flux, and backing materials. These materials are distributed from a central warehouse to sub-warehouses at various workstations via fixed-point delivery. There are 12 welding material distribution points within the shipyard, distributed across ten locations: the Straight Center, Component Platform 1, Component Platform 2, General Assembly Building 1, Control Center, Machining 36m, Curved Surface 45m, Platform 1, Platform 8, and Platform 9B. The welding material warehouse has three transport vehicles with a carrying capacity of 8 tons. The data in the table represents information for each distribution point at a specific time. The first three columns list the names of all warehouses. The last five columns show the current inventory of welding rods, welding wires, carbon rods, flux, and backing materials in each sub-warehouse when the total inventory reaches 5 units, measured in units of 100 kg.

[0054]

[0055] The initial scheduling plan is as follows Figure 4 As shown, this involves two routes and two vehicles. The scheduling order for vehicle 1 is: Main Warehouse → s1 → s2 → s4 → s5 → Main Warehouse. The scheduling order for vehicle 2 is: Main Warehouse → s6 → s0 → s9 → s8 → s7 → s3 → Main Warehouse. During the process of assigning vehicle 1 to branch warehouse 2, a type 2 interruption was encountered. Based on the proposed algorithm, the route was replanned using strategy 1, and the result is as follows. Figure 5 As shown. After replanning, the route for vehicle 1 is: Main Warehouse → s1 → s0 → s9 → s8 → s7 → s3 → Main Warehouse. The route for vehicle 2 remains unchanged: Main Warehouse → s6 → s2 → s5 → s4 → Main Warehouse.

[0056] Example 2: Please refer to Figure 1 This invention provides a technical solution: a method for optimizing vehicle routes in a large manufacturing industry based on adaptive large neighborhood search and inventory shipyards. This method addresses the scheduling problem of production logistics vehicles after interference, specifically when vehicles encounter interference while delivering materials to workstations, affecting their original routes. A series of strategies are employed to restore or optimize these routes to meet the production needs of the workstations. Based on the inventory routing problem, a mathematical model is first proposed to minimize vehicle and delivery costs in the warehouse, finding the optimal delivery route. This route is then applied to the interference recovery model. Furthermore, maximum inventory constraints, safety stock constraints, and vehicle capacity constraints are considered. A hybrid collaborative framework is proposed based on adaptive large neighborhood search to solve these problems. In this framework, warehouse demand is determined by analyzing suitable initial inventory for delivery. An algorithm for encoding and generating initial solutions based on inventory demand and distance is proposed to obtain feasible solutions. Then, through multiple iterations of adaptive large neighborhood search, the optimal vehicle route is obtained. Based on the location of the interference, the optimal route is split and combined, and after multiple iterations, the optimal solution under interference conditions is obtained.

[0057] Due to the limited capacity of the vehicles, the demand for sub-warehouse b cannot be met. pc That is, the distribution quantity of all materials in warehouse i is set to the optimal value, therefore let b pc It is a variable that can be adjusted to approach the optimal inventory level. When the inventory of warehouse material c reaches the inventory warning level sw... c At that time, vehicles originate from the distribution center warehouse and deliver identically packaged goods (c) to the sub-warehouse (p). The quantity delivered is the number of packages, and the space occupied is v. Consider the transport vehicles. The transportation network consists of warehouse P and its sub-warehouses, where all geographical locations (coordinates) are known. The consumption rate of materials c in each sub-warehouse is the same per unit time, denoted as d. pc When developing a delivery plan, the consumption of materials by each warehouse during the delivery process needs to be considered, with the goal of minimizing the delivery route distance and the number of vehicles. Assume there are H types of disturbance events, let H = {1, 2, ..., H}. When a delivery vehicle encounters a disturbance event h ∈ H, determine the type of h, adopt the corresponding strategy, and then decide whether to replan the route. If replanning is necessary, keep the already traveled routes unchanged and only change the untraveled routes. If the inventory of material c in warehouse p is lower than the safety stock value, the difference below the safety stock value is used as the delivery penalty. The disturbance recovery model aims to minimize the path cost, vehicle cost, and delay cost after replanning. The corresponding mathematical model is as follows:

[0058] If no disruptive events occur, only delivery route costs and vehicle costs need to be considered. Within one cycle, a mathematical model and relevant constraints are established, followed by a table explaining the corresponding symbols:

[0059]

[0060] A mathematical model is constructed based on the minimization of vehicle cost and delivery cost, where the first term is the route cost and the second term is the vehicle cost.

[0061] Constraints:

[0062]

[0063] The two formulas above indicate that each sub-repository must be accessed at least once.

[0064]

[0065] The formula indicates that for each sub-warehouse, the number of vehicles entering that point is equal to the number of vehicles leaving that point.

[0066]

[0067] The above two items represent the number of times each vehicle departs from and returns from the main warehouse.

[0068]

[0069] The formula states that the total demand of all warehouses does not exceed the capacity of the vehicles.

[0070]

[0071] The formula states that the number of vehicles assigned to delivery tasks cannot exceed the total number of vehicles, in order to ensure that there are always a certain number of vehicles available as backups.

[0072]

[0073] The two formulas above represent the safety stock constraint and the maximum inventory constraint for any given time period, respectively. The left side of the inequality represents the inventory level when delivery vehicles arrive at the warehouse, which is equal to the initial inventory level minus the consumption level plus the quantity of materials delivered.

[0074]

[0075] The formula ensures that the vehicle's cumulative driving distance does not exceed its maximum driving distance.

[0076] The established mathematical model of the interference event:

[0077]

[0078] In the objective function, Δl i,r This refers to the delay cost that may occur to warehouse i due to a disruption event encountered by delivery vehicle r. This is a constraint imposed on vehicles reaching the warehouse penalty value. VV or ≤V rt ≤V indicates that the capacity of the new vehicles does not exceed the total capacity of the vehicles. This indicates that a vehicle in transit departs from its current location and either goes to an undelivered warehouse or returns directly to the central warehouse. If a transport malfunctions, the vehicle is assumed to return to the central warehouse, and due to the transport vehicle r... t It needs repair and therefore can no longer be used for delivering supplies. This indicates the location from which a newly dispatched vehicle departs from the main depot to the location where the interference event occurred. This indicates that all vehicles must return to the parking lot. left +m new ≤m indicates that the number of newly dispatched vehicles and the total number of vehicles in transit do not exceed the total number of vehicles. The objective function minimizes the total cost and optimizes warehouse performance after a disruptive event occurs.

[0079] Shipyard production cannot be halted, therefore delivery delays cannot exceed safety stock levels. For the above delivery problem to be meaningful, the delay cost Δl for warehouse i is... i Inequalities also need to be satisfied.

[0080] 0≤Δl i,r ≤ls

[0081] At the same time, assuming sufficient vehicles, the initial inventory level of each warehouse must satisfy an inequality for production to operate normally; that is, the delivery vehicles require each warehouse to have sufficient inventory. During their shift, they deliver supplies.

[0082]

[0083] The parameters are explained in the table below:

[0084]

[0085]

[0086] 3. Determine the optimal warehouse early warning value:

[0087] During warehouse operations, the inventory level of the warehouse that reaches the warning threshold is always less than or equal to the inventory level of other warehouses. This means that the delivery volume of the warehouse that reaches the warning threshold is always greater than or equal to the delivery volume of other warehouses. To ensure the delivery tasks are meaningfully completed, let's assume the delivery volume of the warehouse that reaches the warning threshold is b. w The delivery volume of each warehouse should meet b i ≤b w In the design, make b i Expand to b w If we want the delivery vehicles to operate at full capacity as much as possible, then b1 + b2 + ... + b i ≥V, i.e., b w +b w +…+b w If V ≥ 1, and distance costs are not considered, then at least one vehicle is needed to fulfill the delivery needs of n warehouses, then nb w ≥V, b w ≥V / n, therefore the warehouse warning value needs to be set to I. w,i ≤us-V / n. If all vehicles are involved in delivery, and the total amount of goods delivered cannot exceed the total capacity of all vehicles, then the following condition must be met: b1+b2+…+b i ≤mV. Similarly, the delivery volume for each warehouse is increased to b. w At this time there is nb w ≤mV, the warehouse warning value needs to be set to I. w,i≥us-mV / n. The inventory warning value range to ensure the normal operation of the warehouse is [us-mV / n, us-V / n]. The initial inventory needs to meet the following conditions. Therefore, if The lower limit of the inventory warning value is set as follows: Otherwise, set to

[0088] 4. Develop corresponding strategies for different types of interference, and then determine the recovery scheduling scheme: the strategies include a first strategy and a second strategy; wherein:

[0089] Strategy 1: Vehicle self-rescue, i.e., self-rescue by the vehicle during transportation. When a transport vehicle encounters a disruptive event and does not receive or need to receive assistance from a replacement vehicle, it copes with the disruptive event by changing its own route. Strategy 1 is suitable for disruptive events that have a minor impact on vehicle transportation, such as when a vehicle breaks down and needs to wait, in which case the inability to change the route in time leads to a larger overall delay. Applying Strategy 1 to disruptive events in Category 2, the location of the disruptive event is assumed to be a temporary point P. t Remove locations that have already been delivered and replan delivery routes to minimize costs and warehousing efficiency.

[0090] Second Strategy: Dispatch New Vehicles. When encountering disruption events, if vehicles are waiting for too long or there is no reasonable new route to the un-dispatched warehouse (e.g., the route is too long or too costly), a new vehicle is needed to serve the un-dispatched warehouse. Applying this second strategy to disruption events in Category 1, since the total warehouse inventory is limited, it is necessary to pick up the materials from the vehicle experiencing the disruption event. The location of the malfunction is designated as a temporary location, and the new vehicle re-plans its route from this temporary location.

[0091] Both strategies mentioned above increase scheduling costs. The first strategy increases delivery transportation costs and delay costs; the second strategy increases delivery vehicle costs, transportation costs, and delay costs. Therefore, for different disruptive events, it is necessary to select an appropriate strategy to minimize delivery costs and delay costs.

[0092] 5. A hybrid collaborative framework based on large neighborhood search was designed to solve the delivery model of vehicles encountering interference events, including a solution generation algorithm, an algorithm for determining warehouse replenishment materials based on performance value adjustment strategies, and large neighborhood removal and repair operators based on this problem:

[0093] To ensure the adaptive large neighborhood search algorithm runs quickly and efficiently and yields excellent solutions, a suitable initial solution needs to be designed based on the distance between warehouses and their current inventory. Generally, warehouses with smaller initial inventory differences are less likely to be delivered sequentially by the same vehicle, while warehouses with larger initial inventory differences and smaller distance differences are more likely to be delivered sequentially by the same vehicle. A warehouse can be considered the next delivery target if the delivery vehicle has served a warehouse and the delivery time to the next warehouse is less than the time it takes for materials to reach safety stock. A greedy rule is used to select the next delivery warehouse; that is, before a warehouse reaches its safety stock, the nearest warehouse is prioritized as the next delivery point. The solution is encoded in row-major form. For example, when there are 10 warehouses, [2,3,1,6,5,4,8,9,10,7] is an encoding without the total inventory. If the total warehouse is included, the code is [0,2,3,1,0,6,5,4,8,0,9,10,7,0], where 0 represents the total warehouse and [0,2,3,1,0] represents a delivery route.

[0094] As the vehicle delivery process shows, when a warehouse issues a warning due to reduced supplies, vehicles need to be dispatched for delivery. Differences in production processes at various warehouse workstations mean that warehouses do not all issue delivery warnings at the same time. However, to maximize vehicle loading rates, all warehouses are replenished with appropriate amounts of supplies. The amount of supplies warehouse i needs to replenish is b. i A vehicle assigned to the optimal delivery route can deliver a quantity equal to the sum of the materials needed by several warehouses, i.e., b1 + b1 + ... + b i The following situations may occur during the delivery process:

[0095] Total amount of materials But b1+b1+…+b i In cases where the value is greater than V, rescheduling to use another vehicle would increase delivery costs. Therefore, it's necessary to adjust the delivery volume, appropriately reducing the amount of goods in the delivery warehouse, so that the total amount of goods is b1 + b1 + ... + b i =V.

[0096] The total quantity of materials is b1 + b1 + ... + b i In the case where V = 0, there is no need to adjust the warehouse's delivery volume.

[0097] Total amount of materials But b1+b1+…+b i In the case of <V, adding a warehouse would far exceed the vehicle's capacity for delivery. Not increasing the delivery volume would increase the vehicle's empty load rate. Therefore, it's necessary to adjust the delivery volume by appropriately increasing the amount of goods delivered from the warehouse, so that the total amount of goods is b1 + b1 + ... + b i =V.

[0098] This inventory cost performance-based strategy allocates delivery volumes to warehouses, enabling them to operate normally and achieve optimal inventory performance without excessively increasing delivery costs. This algorithm is based on the greedy logic often employed by warehouse managers in their decision-making: the farther the distribution center, the more materials are delivered. Since warehousing performance values ​​are only related to the target inventory level and delivery volume, only these two factors are considered when developing the algorithm.

[0099] After generating the initial solution, the optimal solution to the problem needs to be found. An adaptive large neighborhood search algorithm is used to find the optimal solution. This algorithm expands the search space by designing multiple sets of destruction and repair operators, improves the current solution, and scores or weights the operators. Operators with better performance receive higher scores or weights. In each iteration, the destruction and repair operators are selected and their weights adjusted based on past performance. An efficient combination method is used to improve the algorithm's optimization ability, thereby finding the optimal solution.

[0100] (1) Removal Operator. Based on this problem, five destruction operators are introduced:

[0101] ① Worst path removal operator: Select the two paths with the highest cost, then remove one warehouse from each of the two paths, and repeat until all warehouses on both paths are removed.

[0102] ② Delivery removal operator: When delivering to each warehouse by vehicle, there exists b1 + b1 + ... + b i <V, and b1+b1+…+b i When the difference between the warehouse and V is small, the warehouse with the smallest delivery volume needs to be removed and replaced with a warehouse with a slightly larger delivery volume to maximize vehicle utilization. Simultaneously, there exists b1 + b1 + ... + b i >V, and b1+b1+…+b i If the difference between V and V is small, then the warehouse with the largest delivery volume needs to be removed and a warehouse with a slightly smaller delivery volume should be inserted so that the total delivery volume does not exceed the vehicle's capacity by too much.

[0103] ③ Neighborhood Removal Operator: Warehouses with similar current inventory levels are more likely to cause delivery delays. Therefore, it is necessary to determine the warehouses with the current inventory levels along each path and randomly remove warehouses with smaller inventory differences. If multiple warehouses exist, a strategy of randomly selecting one is adopted.

[0104] ④ Location Removal Operator: Using the total warehouse as the origin, divide all warehouses into four regions, and check the number of warehouses belonging to each region along the path. Based on the number of warehouses belonging to a region, set the probability of removing a warehouse, and then randomly remove warehouses.

[0105] ⑤ Shortest path removal operator: Determine if there exists a shortest path among all paths, and if the total number of repositories contained in that path does not exceed the total number of paths, then remove that path.

[0106] Using the removal method described above, all removed repositories are placed into a set. Then, by proposing an existing quantity insertion method, combined with a repair method based on greedy insertion and regretful insertion proposed by Ropke & Pisinger, the repositories in the set are inserted into the appropriate paths.

[0107] ① Existing quantity insertion: Compare the existing quantity of removed warehouse materials, and prioritize the previously destroyed warehouses according to the order of existing quantity, from smallest to largest, and randomly insert them into any line in the way the initial solution was generated.

[0108] ② Greedy insertion: Greedy insertion involves sequentially placing the nodes to be inserted into all positions in the line and recording the position with the best cost change. At this time, it is not necessary to consider the total amount of delivered materials, similar to the delivery removal operator.

[0109] ③ Regretful Insertion: Regretful insertion is the difference between the cost calculated after inserting a warehouse at the optimal position and the cost calculated after inserting it at the second-best position. A higher regret value for a warehouse indicates that if it is not inserted at the optimal position now, the efficiency of inserting it at other positions later will be greatly reduced. The process iteratively calculates the regret values ​​of n warehouses that need to be inserted, finds the node with the highest regret value among the current n warehouses, inserts it into the line, and updates the cost. This process is repeated to calculate the regret values ​​of n-1 warehouses that need to be inserted, finds the warehouse with the highest regret value among the n-1 warehouses that need to be inserted, and inserts it. This process is repeated until all removed warehouses are repaired.

[0110] (2) Dynamically adjust weights and select

[0111] The algorithm iterates by selecting, adjusting, removing, and repairing operators based on operator weights and a roulette wheel selection method. The roulette wheel method assigns weights to each operator and calculates the probability of its occurrence to ultimately select the operator. The operator selection probability is determined by a formula. For the removal operator, k=5; for the repair operator, k=3. Record the performance of each removal and repair operator, and dynamically adjust the weights w. i Initially, each operator is assigned the same weight, and the weights are continuously updated based on the quality of the results obtained. The weights are updated using formula (15), where λ is the weight update coefficient, set to 0.4; s i μ is the score of the operator. i denoted as the number of times the operator appears.

[0112]

[0113] When the iteration is complete, scores s for the destruction and repair methods used in the last iteration are calculated using different rules. i If the result obtained is the globally best result, then s i =30; if the obtained result is better than the original result, but worse than the optimal result, then s i =20; if the result obtained is worse than the original result, then s i =10. Finally, using the Metropolis criterion in the simulated annealing algorithm, the superior and inferior solutions are obtained under the selected probabilities.

[0114] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

[0115] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for optimizing vehicle routes in a shipyard based on adaptive large neighborhood search and inventory shipyard, characterized in that, Includes the following steps: A mathematical model is established based on initial inventory and demand, and a mathematical model is established within a cycle based on delivery route costs and vehicle costs. The mathematical model includes: The objective function is to minimize the total cost, which includes delivery route costs and vehicle costs. The constraints are vehicle capacity, warehouse demand, number of vehicles, and routes, among which: Vehicle capacity constraint: Each vehicle's delivery volume cannot exceed its maximum capacity; Warehouse demand constraint: The demand for each warehouse must be met; Vehicle number constraint: The number of vehicles used cannot exceed the total number of available vehicles; By combining vehicle capacity and the distance between warehouses, a feasible initial delivery route is generated; the logic for obtaining the initial delivery route is as follows: Based on a hybrid collaboration framework, warehouse demand is determined by analyzing the initial inventory suitable for warehouse delivery. An algorithm for encoding and generating initial solutions based on inventory demand and distance is proposed to obtain feasible solutions; The optimal vehicle route is obtained through multiple iterations of adaptive large neighborhood search. The initial delivery route is iteratively optimized using an adaptive large neighborhood search algorithm to obtain the optimal delivery route. Based on the maximum inventory constraint, safety stock constraint, and vehicle capacity constraint, the optimal delivery route is applied to the disturbance recovery model for decomposition and combination. After multiple iterations, the optimal solution under the disturbance state is obtained. Mathematical models for interference events include: When a disruptive event occurs, the objective function is to minimize the total cost, which includes delivery route costs, vehicle costs, and delay costs. The constraints are: interference events, vehicle self-rescue, dispatching new vehicles, and safety stock. Interference event constraints: Consider the impact of interference events on delivery routes; Vehicle self-rescue restraint: When a vehicle encounters interference, it can self-rescue by changing its driving route; Additional vehicle deployment constraint: When necessary, additional vehicles should be deployed to serve warehouses that have not been dispatched. Safety stock constraint: Ensure that the inventory level in each warehouse is not lower than the safety stock value; Based on the type and impact of the interference, corresponding recovery strategies are adopted to determine the recovery and dispatch plan; strategies include vehicle self-rescue and dispatching additional vehicles.

2. The method for optimizing vehicle routes in a shipyard based on adaptive large neighborhood search and inventory shipyards according to claim 1, characterized in that, The warehouse replenishment quantity and vehicle capacity are determined based on the optimal performance strategy to determine the warehouse distribution combination; The delivery combination of warehouses is adjusted based on their distance, and the order of warehouses is determined based on the initial inventory of the warehouses to integrate vehicle delivery routes; Determine if any disruptions have occurred to the delivery vehicle; If no disruptions occur to the delivery vehicles, the delivery route will be replanned to minimize the total cost. If a disruption event occurs to delivery vehicles, the type of disruption event is determined, and the location of each delivery vehicle at the time of the disruption event is identified. In this way, appropriate recovery strategies are adopted to handle the situation and determine the recovery scheduling plan.

3. The method for optimizing vehicle routes in a shipyard based on adaptive large neighborhood search and inventory shipyards according to claim 2, characterized in that, Develop an optimal warehouse early warning value, mark the optimal warehouse early warning value as the preset early warning value, and develop different recovery strategies based on the preset early warning value. The recovery strategies include a first strategy and a second strategy. If the warehouse warning value is lower than the preset warning value, the first strategy is selected: vehicle self-rescue, which involves changing the delivery route. If the warehouse warning value is greater than the preset warning value, the second strategy is selected, and new vehicles are dispatched. This is suitable for situations where the waiting time is too long or there is no reasonable new route.

4. The method for optimizing vehicle routes in a shipyard based on adaptive large neighborhood search and inventory shipyards according to claim 3, characterized in that, The hybrid collaborative framework based on large neighborhood search includes: a route generation algorithm, an algorithm for determining warehouse replenishment materials based on performance value adjustment strategies, and large neighborhood removal and repair operators based on this optimization problem.

Images