Single-track double-car scheduling method for seamless stainless steel pipe cold rolling process

Abstract

The application discloses a single-track double-car scheduling method for a seamless stainless steel pipe cold rolling process, and comprises the following steps: S10, setting an RVG transport path, wherein a running path of a first trolley RGV1 is as follows: a raw material warehouse station S n As a starting station, raw materials are first loaded, and then the station S n After driving to a requested machine tool station, the raw materials are unloaded, and then the trolley continues to drive to the raw material warehouse station S n of the starting point, and stops to wait for execution of a next transport task; a running path of a second trolley RGV2 is as follows: starting from a second station S2, driving to the requested machine tool station, loading finished products, then driving to a warehouse station S1, unloading the finished products, then continuing to drive to the second station S2 of the starting point, and stopping to wait for execution of the next transport task; S20, triggering a delivery demand according to a machine tool working state; and S30, performing RGV dynamic scheduling.

Description

Technical Field

[0001] This invention belongs to the field of scheduling technology, specifically relating to a single-track dual-car scheduling method for the cold rolling process of seamless stainless steel pipes. Background Technology

[0002] Linear reciprocating RGV (Rail Guided Vehicle) scheduling systems are divided into single RGV mode and dual RGV mode based on different operating modes. In single RGV transport, a single RGV trolley is commonly used for material handling in production, moving back and forth on a straight track. This method is simple to schedule, requiring only consideration of the single vehicle's route and task allocation. However, its efficiency is low. When faced with multiple machine tools and frequent loading and unloading demands for raw materials and finished products, the limited transport capacity of a single RGV trolley can easily lead to machine tools waiting for raw materials or finished products accumulating, affecting the overall production schedule. To improve efficiency, dual RGV transport was adopted, but initially, there was no reasonable scheduling plan to handle operational conflicts. With two RGVs running on the same track, spatial and temporal conflicts are likely to occur. For example, the two RGVs may have overlapping routes and arrive at the same area within the same time interval, without an effective avoidance mechanism, leading to transport delays or even collisions, thus jeopardizing smooth production.

[0003] In terms of scheduling algorithms, previous methods simply sorted tasks and dispatched vehicles according to the order of task triggering or fixed rules. This approach did not fully consider the dynamic changes in machine tool operating status, the urgency of tasks, and the real-time position and status of the RGV (Automated Guided Vehicle) carts. In actual production, the raw material processing time and the time for the finished product buffer to fill on the machine tool are constantly changing. Simple task sorting and scheduling cannot adapt to this dynamism, resulting in suboptimal scheduling schemes that affect production efficiency and resource utilization. Summary of the Invention

[0004] In view of the above-mentioned problems, the present invention provides a single-track dual-vehicle scheduling method for the cold rolling process of seamless stainless steel pipes. By using the dual judgment conditions of spatial conflict and temporal conflict, combined with targeted avoidance strategies, the running conflict of the two RGVs on the straight track can be accurately avoided, the risk of equipment collision can be eliminated, the transportation process can be ensured to be stable and uninterrupted, and the safety and smoothness of the production loading and unloading links can be guaranteed.

[0005] To solve the above-mentioned technical problems, the present invention adopts the following technical solution: A single-rail dual-car scheduling method for the cold rolling process of seamless stainless steel tubes includes the following steps: S10, Set the RVG transport route: Along the transport route, set up warehouse station S1, second station S2, and multiple machine tool stations S3 to S4 in sequence. n-1 and raw material storage site S nThe first trolley, RGV1, is used to transport raw materials to various machine tools, and the second trolley, RGV2, is used to transport finished products to the warehouse. The operating path of the first trolley, RGV1, is as follows: Raw material warehouse station S n As the starting point, first load the raw materials, then proceed to station S. n After unloading the raw materials at the requested machine tool station, the vehicle continues to the raw material warehouse station S at the starting point. n Stop and wait to execute the next transport task; the running path of the second trolley RGV2 is as follows: starting from the second station S2, travel to the requesting machine tool station, load the finished product, then travel to the warehouse station S1, unload the finished product, continue to travel to the second station S2 at the starting point, stop and wait to execute the next transport task; S20: Trigger transportation requests based on machine tool operating status: trigger a raw material transportation request when the machine tool is short of raw materials; trigger a finished product transportation request when the machine tool's finished product buffer is full. This represents the remaining time for machine tool i to finish processing the raw material at time t. This represents the remaining time before machine tool i's finished product buffer is full at time t. and It decreases with time t; when When =0, it indicates that the machine tool is short of raw materials, triggering a raw material loading request, which is recorded as . ;when When the timer value is 0, the finished product buffer of the machine tool is full, triggering a finished product unloading request. The machine tool continues to operate. When machine tool i is replenished with raw materials at time t, then... When machine tool i removes the finished product at time t, then ; S30, perform RGV dynamic scheduling: At time t, the remaining time for each machine tool to finish processing raw materials and the remaining time for the finished product buffer to be full are respectively... , , , ... , Select the k smallest remaining time intervals. The raw material or finished product loading / unloading requests that will be triggered after the k remaining time intervals expire constitute the transportation task candidate set R. For the task candidate set R, it is divided into R based on the RGV of the task execution. 1 and R 2 Two candidate subsets of tasks are given, representing the candidate subsets of raw material transportation tasks and the candidate subsets of finished product transportation tasks, respectively, and denoted as R. 1 and R 2 The number of elements are k1 and k2 respectively, and k1 + k2 = k; R 1 and R 2Any arrangement of tasks in the candidate set constitutes an RGV scheduling scheme, where RGV1 and RGV2 are respectively arranged according to R... 1 and R 2 The task order of the candidate set is dynamically scheduled using RGV.

[0006] In one possible implementation, dynamic RGV scheduling in S30 includes: S301, if k1≠0 and k2≠0, then use a genetic algorithm to find R. 1 and R 2 Optimal task arrangement order, and based on R 1 and R 2 If the task is scheduled in the correct order, dispatch RGV1 and RGV2 vehicles to S304; otherwise, proceed to S302. S302, if k1=0 and k2≠0, then according to R 2 The tasks are sorted according to their trigger times to determine the task arrangement, and then based on R... 2 If the task is scheduled in the correct order, proceed to S304; otherwise, proceed to S303. S303, if k1≠0 and k2=0, then according to R 1 The tasks are sorted according to their trigger times to determine the task arrangement, and then based on R... 1 The task is scheduled to dispatch RGV1 trolley, then transfer to S304; S304, after scheduling m RGV vehicles, dynamically update the task candidate set and recalculate the task candidate set R.

[0007] In one possible implementation, a genetic algorithm is used in S301 to find R. 1 and R 2 The optimal task order includes: S3011 employs a direct permutation encoding method based on task execution order, where each chromosome represents a complete scheduling scheme, determined by R. 1 and R 2 The arrangement and composition of task numbers; S3012, the initial population is constructed using a random permutation method, for R. 1 and R 2 The task generates x distinct task permutations independently, each permutation being a random permutation of the task numbers, resulting in a population size of x. 2 ; S3013, an evaluation system is constructed based on the negative correlation between fitness and total completion time makespan. The fitness function is defined as the reciprocal of the total completion time: fitness = 1 / makespan; S3014, for a given R 1 and R2 The task order is arranged, and the task scheduling process is simulated based on the current states of RGV1 and RGV2 vehicles to calculate the total completion time.

[0008] In one possible implementation, the RGV state is set as follows: The RGV state is described by a six-tuple, which is null when idle. The six-tuple includes the task ID, the task machine tool docking station ID, the current task start time, the time of arrival at the task machine tool docking station ID, the time of departure from the task machine tool docking station ID, and the time of return to the starting station of the travel path. The task ID is the task number for either the raw material transport task or the finished product transport task; the machine tool docking station ID is the machine tool docking station number that the current trolley needs to reach for the raw material transport task or the finished product transport task; the start time of the current task refers to the time when the trolley departs from the starting station of the travel path to start executing the raw material transport task or the finished product transport task, using the global system time; the time when the trolley arrives at the machine tool docking station ID is the system time when the trolley arrives at the designated machine tool docking station for the current raw material transport task or the finished product transport task; the time when the trolley leaves the machine tool docking station ID is the system time when the trolley arrives at the designated machine tool docking station, unloads raw materials or loads finished products and leaves the station; the time when the trolley returns to the starting station of the travel path is the system time when the trolley returns to the starting station after leaving the machine tool docking station, expressing the final system time when the trolley completes the current raw material transport task or the finished product transport task.

[0009] In one possible implementation, the process of calculating the total completion time is as follows: S30141, Obtain R 1 and R 2 The system time for triggering all tasks is determined. The current system time is obtained and denoted as T1. The initial value of the simulation scheduler clock T is set to T1. The states V1 and V2 of RGV1 and RGV2 are obtained. V1= ; V2= ; S30142, If R 1 and R 2 If all queues are empty, proceed to S30147; S30143, if RGV1 is idle and RGV2 is idle, and R 1 and R 2 If the queue is not empty, then start from R. 1 and R 2 Take one task from the head of each queue, namely R 1 With R 2 Compare R 1 With R2 The task trigger time, if R 1 If the time is earlier, then first determine T and R. 1 The size of the task trigger time, if R 1 If the task trigger time is longer, then T is updated to R. 1 Task trigger time; secondly, the RGV1 car simulates the execution of R... 1 Task: Update the status of RGV1 car V1: V1= ,in It is R 1 The task number; It is R 1 Machine tool docking station number; The value is T; The value is + ; The value is +Unloading time of raw materials; The value is + Third, when RGV2 is idle, its state V2 is set to Null; finally, R... 1 Task from R 1 Delete from the queue; If R 2 If the time is earlier, then first determine T and R. 2 The size of the task trigger time, if R 2 If the task trigger time is longer, then T is updated to R. 2 Task trigger time; secondly, the RGV2 car simulates the execution of R... 2 Task: Update the status of RGV2 car V2; V2= ,in It is R 2 The task number; It is R 2 Machine tool docking station number; The value is T; The value is +Time to load raw materials at the starting point of the trolley+ ; The value is + Loading time for finished products; The value is +Uninstallation time+ + Third, when RGV1 is idle, its state V1 is set to Null; finally, R... 2 Task from R 2 Delete from the queue; S30144, if RGV1 is idle and R1 If the queue is not empty, then start from R. 1 Take a task R from the head of the queue. 1 If R 2 The queue is empty or R 1 The task trigger time is less than V2 If the value is T, then T is updated to R. 1 The task trigger time is set; otherwise, the RGV2 car completes its transportation task, returns to idle, updates the RGV2 car status V2 to Null, and proceeds to S30142. RGV1 car simulates execution R 1 Task: Update the status of RGV1 car V1= ,in It is R 1 The task number; It is R 1 Machine tool docking station number; The value is T; The value is + ; The value is +Unloading time of raw materials; The value is + Finally, R 1 Task from R 1 Delete from the queue; S30145, if RGV2 trolley is idle and R 2 If the queue is not empty, then start from R. 2 Take a task R from the head of the queue. 2 If R 1 The queue is empty or R 2 The task trigger time is less than V1 If the value is T, then T is updated to R. 2 The task trigger time is set; otherwise, the RGV1 car completes its transportation task, returns to idle, updates the RGV1 car status V1 to Null, and proceeds to S30142. RGV2 car simulates execution R 2 Task: Update the status of RGV2 car V2= ,in It is R 2 The task number; It is R 2 Machine tool docking station number; The value is T; The value is +Time to load raw materials at the starting point of the trolley+ ; The value is + Loading time for finished products; The value is +Uninstallation time+ + Finally, R 2 Task from R 2 Delete from the queue; S30146, If neither the state V1 of RGV1 nor the state V2 of RGV2 is null, then determine whether there is a conflict between the two RGV vehicles during operation, so as to decide whether one vehicle needs to give way and wait for the other two vehicles. S30147, if neither V1 nor V2 of the vehicle is null, then compare... and ,when When the value of T is smaller, it is set to... And RGV1 car is idle and V1 is set to Null; when When the value of T is smaller, it is set to... Furthermore, RGV2 is idle and V2 is set to null; if the vehicle state V1 is not null and V2 is null, then the value of T is set to... Furthermore, if RGV1 is idle and V1 is set to null; if the vehicle state V2 is not null and V1 is null, then the value of T is set to... Furthermore, RGV2 is set to idle and V2 is set to null; the total completion time makespan is returned as T. T1.

[0010] In one possible implementation, in S30146, the conditions under which the two RGV vehicles conflict during operation include: spatial conflict conditions, and the state V1 of RGV1 vehicle... Less than RGV2, the car's state V2 This means that the travel paths of the two cars overlap; under the condition of temporal conflict and spatial conflict, the RGV1 car starts from the station. Arrive at the station and on the site The time interval Q1 for unloading raw materials and leaving is [ , The RGV2 car departs from the station. Arrive at the station and on the site The time interval Q2 for loading and leaving the finished product is [ , A time conflict refers to an overlap between intervals Q1 and Q2. If two RGV vehicles simultaneously meet both spatial and temporal conflict conditions, then a conflict exists between the two RGV vehicles during operation. The strategy for handling the conflict is... Less than Then the RGV2 car is numbered Station 1 is waiting to be cleared, with a waiting time w. And the RGV2 car status V2 is updated to Otherwise, the RGV1 car will be numbered. Station 1 is waiting to be cleared, with a waiting time w. And the RGV1 car status V1 is updated to .

[0011] The present invention has the following beneficial effects: (1) By clearly defining the dual judgment conditions of spatial conflict and temporal conflict (the station numbers of the two vehicles overlap in space, and the time intervals of the two vehicles' travel and parking overlap in time), combined with targeted avoidance strategies (assigning one vehicle to wait at a specific station based on the departure time of the two vehicles), the running conflict of the two RGVs on the straight track can be accurately avoided, the risk of equipment collision can be eliminated, the transportation process can be ensured to be stable and uninterrupted, and the safety and smoothness of the production loading and unloading links can be guaranteed.

[0012] (2) Based on the real-time working status of the machine tool (via and The remaining processing time of raw materials and the remaining time of finished product buffer are monitored in real time, respectively, and a candidate set R for transportation tasks is generated. The candidate set of tasks is dynamically updated after each m RGV scheduling, so that the scheduling scheme is always synchronized with the real-time demand of machine tools for raw material shortage and buffer fullness. This completely avoids the lag problem of traditional static scheduling being unable to respond to dynamic changes in production and ensures continuous operation of machine tools.

[0013] (3) With “minimizing the total completion time (makespan)” as the core objective, the genetic algorithm is used to optimize the arrangement of R1 (raw material transportation task) and R2 (finished product transportation task): chromosomes are constructed by task order encoding, diverse initial populations are generated by random arrangement, and a fitness evaluation function is established to select high-quality solutions with “1 / makespan” as the fitness function. Combined with tournament selection, sequential crossover, exchange mutation and elite retention strategies, the optimal task arrangement is searched efficiently, significantly shortening the total time of the overall transportation task and improving scheduling efficiency.

[0014] (4) For different combinations of the number of candidate task subsets R1 and R2 (k1≠0 and k2≠0, k1=0 and k2≠0, k1≠0 and k2=0), a differentiated scheduling logic is designed: when two task sets coexist, a genetic algorithm is used for optimization; when a single task set exists, it is sorted according to the trigger time. This not only covers all possible task scenarios in production and ensures the universality of the solution, but also allows for flexible adjustment of the scheduling strategy according to the actual task type, avoiding the waste of efficiency caused by "one-size-fits-all" scheduling.

[0015] (5) The status of the RGV vehicle is accurately quantified by using a six-tuple (task ID, task machine tool docking station ID, start time of the current task of the vehicle, time when the vehicle arrives at the task machine tool docking station ID, time when the vehicle leaves the task machine tool docking station ID, and time when the vehicle returns to the starting station of the driving path). At the same time, the time calculation rules for the transportation of raw materials or finished products are clarified, and the correction of avoidance waiting time is included. This ensures that the time nodes and vehicle status of the scheduling plan have clear quantitative basis, greatly improves the scheduling accuracy, and can be directly implemented to avoid the disconnect between theory and practice. Attached Figure Description

[0016] Figure 1 This is a flowchart illustrating the steps of a single-track dual-car scheduling method for the cold rolling process of seamless stainless steel tubes according to an embodiment of the present invention. Figure 2 This is a schematic diagram of a linear reciprocating double RGV production loading and unloading station in an embodiment of the present invention. Detailed Implementation

[0017] 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, not all, of the embodiments of the present invention. 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.

[0018] Reference Figure 1 The diagram shows a flowchart of a single-rail dual-car scheduling method for the cold rolling process of seamless stainless steel tubes according to an embodiment of the present invention, including the following steps: S10, Set the RVG transport route: (e.g., ...) Figure 2 The diagram shows a linear reciprocating double RGV production loading and unloading station in an embodiment of the present invention. During production, raw material loading (transporting from the raw material warehouse to the machine tool) and finished product unloading (transporting from the machine tool to the warehouse) are carried out using linear track RGV trolleys. The transport path is sequentially set with warehouse station S1, second station S2, and multiple machine tool stations S3 to S4. n-1 and raw material storage site S n The first trolley, RGV1, is used to transport raw materials to various machine tools, and the second trolley, RGV2, is used to transport finished products to the warehouse. The operating path of the first trolley, RGV1, is as follows: Raw material warehouse station S n As the starting point, first load the raw materials, then proceed to station S. n After unloading the raw materials at the requested machine tool station, the vehicle continues to the raw material warehouse station S at the starting point. nStop and wait to execute the next transport task; the running path of the second trolley RGV2 is as follows: starting from the second station S2, travel to the requesting machine tool station, load the finished product, then travel to the warehouse station S1, unload the finished product, continue to travel to the second station S2 at the starting point, stop and wait to execute the next transport task; S20: Trigger transportation requests based on machine tool operating status: trigger a raw material transportation request when the machine tool is short of raw materials; trigger a finished product transportation request when the machine tool's finished product buffer is full. This represents the remaining time for machine tool i to finish processing the raw material at time t. This represents the remaining time before machine tool i's finished product buffer is full at time t. and It decreases with time t; when When =0, it indicates that the machine tool is short of raw materials, triggering a raw material loading request, which is recorded as . ;when When the timer value is 0, the finished product buffer of the machine tool is full, triggering a finished product unloading request. The machine tool continues to operate. When machine tool i is replenished with raw materials at time t, then... When machine tool i removes the finished product at time t, then ; S30, Perform dynamic RGV scheduling: Due to conflicts arising during the operation of the machine tool's linear reciprocating dual RGVs, a scheduling scheme for the RGVs needs to be formulated based on the machine tool's working status. At time t, the remaining time for each machine tool to complete material processing and the remaining time for the finished product buffer to be full are respectively... , , , ... , Select the k smallest remaining time intervals. The raw material or finished product loading / unloading requests that will be triggered after the k remaining time intervals expire constitute the transportation task candidate set R. For the task candidate set R, it is divided into R based on the RGV of the task execution. 1 and R 2 Two candidate subsets of tasks are given, representing the candidate subsets of raw material transportation tasks and the candidate subsets of finished product transportation tasks, respectively, and denoted as R. 1 and R 2 The number of elements are k1 and k2 respectively, and k1 + k2 = k; R 1 and R 2 Any arrangement of tasks in the candidate set constitutes an RGV scheduling scheme, where RGV1 and RGV2 are respectively arranged according to R... 1 and R 2 The task order of the candidate set is dynamically scheduled using RGV.

[0019] Furthermore, conflicts arise during the operation of the linear reciprocating dual RGV system. Different arrangements of candidate sets R1 and R2 will lead to different completion times for the transported tasks within the candidate sets. Therefore, it is necessary to coordinate the arrangement of tasks in candidate sets R1 and R2 to find the optimal scheduling scheme. The machine tool continuously replenishes raw materials and retrieves finished products, enabling it to operate continuously. Dynamic RGV scheduling in S30 includes: S301, if k1≠0 and k2≠0, then use a genetic algorithm to find R. 1 and R 2 Optimal task arrangement order, and based on R 1 and R 2 If the task is scheduled in the correct order, dispatch RGV1 and RGV2 vehicles to S304; otherwise, proceed to S302. S302, if k1=0 and k2≠0, then according to R 2 The tasks are sorted according to their trigger times to determine the task arrangement, and then based on R... 2 If the task is scheduled in the correct order, proceed to S304; otherwise, proceed to S303. S303, if k1≠0 and k2=0, then according to R 1 The tasks are sorted according to their trigger times to determine the task arrangement, and then based on R... 1 The task is scheduled to dispatch RGV1 trolley, then transfer to S304; S304, after scheduling m RGV vehicles, dynamically update the task candidate set and recalculate the task candidate set R.

[0020] Furthermore, in S301, a genetic algorithm is used to find R. 1 and R 2 The optimal task order includes: S3011 employs a direct permutation encoding method based on task execution order, where each chromosome represents a complete scheduling scheme, determined by R. 1 and R 2 The arrangement and composition of task numbers; S3012, the initial population is constructed using a random permutation method, for R. 1 and R 2 The task generates x distinct task permutations independently, each permutation being a random permutation of the task numbers, resulting in a population size of x. 2 This ensures the diversity of the initial population, laying the foundation for global search, while avoiding premature convergence caused by overly similar initial solutions.

[0021] S3013 constructs an evaluation system based on the negative correlation between fitness and total completion time (makespan). The fitness function is defined as the reciprocal of the total completion time: fitness = 1 / makespan. This ensures that scheduling schemes with shorter total completion times have higher fitness values, thus gaining a greater probability of survival during the selection process. This monotonically decreasing relationship ensures consistency between the optimization objective and the fitness evaluation.

[0022] S3014, for a given R 1 and R 2 The task order is arranged, and the task scheduling process is simulated based on the current states of RGV1 and RGV2 vehicles to calculate the total completion time.

[0023] During the iterative process of the genetic algorithm, tournament selection is used to select superior parent individuals from the current population. Offspring are generated through sequential crossover to inherit superior genes from their parents. Diversity is introduced by crossover (randomly swapping two task positions). Simultaneously, an elite retention strategy is employed to directly replicate the best individual from each generation to the next, ensuring the algorithm converges to a high-quality solution and thus efficiently searching for R0. 1 and R 2 Optimal task order.

[0024] Specifically, in a concrete application example, the RGV state is set as follows: The RGV state is described by a six-tuple, which is null when idle. The six-tuple includes the task ID, the task machine tool docking station ID, the current task start time, the time of arrival at the task machine tool docking station ID, the time of departure from the task machine tool docking station ID, and the time of return to the starting station of the travel path.

[0025] The task ID is the task number for either the raw material transport task or the finished product transport task. The machine tool docking station ID is the machine tool docking station number that the current trolley needs to reach for the raw material transport task or finished product transport task. The trolley's current task start time refers to the start time when the trolley departs from the starting station of the travel path to execute the raw material transport task or finished product transport task, using the global system time. The trolley's arrival time at the task machine tool docking station ID is the system time when the trolley arrives at the designated machine tool docking station for the current raw material transport task or finished product transport task. For the raw material transport task, the initial value of the trolley's arrival time at the task machine tool docking station = trolley's current task start time + trolley's raw material loading time at the starting point + For finished product transportation tasks, the initial time for the trolley to arrive at the task machine tool's stopping point = the start time of the trolley executing the current task + Because of potential conflicts during the operation of the linear reciprocating dual RGVs, if the trolley needs to yield and wait for another two trolleys, the updated time for the trolley to arrive at the task machine tool's stopping point is the initial time plus the yielding waiting time. The time the trolley leaves the task machine tool's stopping point is the system time after the trolley arrives at the designated machine tool's stopping point, unloads raw materials, or loads finished products and leaves the point. For raw material transport tasks, the initial time for the trolley to leave the task machine tool's stopping point = the initial time for the trolley to arrive at the task machine tool's stopping point + the time for unloading raw materials. For finished product transport tasks, the initial time for the trolley to leave the task machine tool's stopping point = the initial time for the trolley to arrive at the task machine tool's stopping point + the time for loading finished products. If the trolley needs to yield and wait for another two trolleys, the updated time for the trolley to leave the task machine tool's stopping point is the initial time plus the yielding waiting time. The time the trolley returns to the starting point of its travel path is the system time after the trolley leaves the task machine tool's stopping point and returns to the starting point, representing the final system time for the trolley to complete the current raw material transport task or finished product transport task. For raw material transportation tasks, the initial time for the trolley to return to the starting station of the travel path = the initial time for the trolley to leave the machine tool's docking station + For the finished product transportation task, the initial time for the trolley to return to the starting station of the travel path = the initial time for the trolley to leave the task machine tool's docking station + the unloading time of the finished product + + If the car needs to yield and wait for another two cars, the update time for the car to return to the starting station of the travel path is the initial time plus the yielding and waiting time.

[0026] Furthermore, the total completion time is calculated in S3014 as follows: S30141, Obtain R 1 and R 2 The system time for triggering all tasks is determined. The current system time is obtained and denoted as T1. The initial value of the simulation scheduler clock T is set to T1. The states V1 and V2 of RGV1 and RGV2 are obtained. V1= ; V2= ; S30142, If R 1 and R 2 If all queues are empty, proceed to S30147; S30143, if RGV1 is idle and RGV2 is idle, and R 1 and R 2 If the queue is not empty, then start from R. 1 and R 2 Take one task from the head of each queue, namely R 1 With R 2 Compare R 1With R 2 The task trigger time, if R 1 If the time is earlier, then first determine T and R. 1 The size of the task trigger time, if R 1 If the task trigger time is longer, then T is updated to R. 1 Task trigger time; secondly, the RGV1 car simulates the execution of R... 1 Task: Update the status of RGV1 car V1: V1= ,in It is R 1 The task number; It is R 1 Machine tool docking station number; The value is T; The value is + ; The value is +Unloading time of raw materials; The value is + Third, when RGV2 is idle, its state V2 is set to Null; finally, R... 1 Task from R 1 Delete from the queue; If R 2 If the time is earlier, then first determine T and R. 2 The size of the task trigger time, if R 2 If the task trigger time is longer, then T is updated to R. 2 Task trigger time; secondly, the RGV2 car simulates the execution of R... 2 Task: Update the status of RGV2 car V2; V2= ,in It is R 2 The task number; It is R 2 Machine tool docking station number; The value is T; The value is +Time to load raw materials at the starting point of the trolley+ ; The value is + Loading time for finished products; The value is +Uninstallation time+ + Third, when RGV1 is idle, its state V1 is set to Null; finally, R... 2 Task from R 2 Delete from the queue; S30144, if RGV1 is idle and R1 If the queue is not empty, then start from R. 1 Take a task R from the head of the queue. 1 If R 2 The queue is empty or R 1 The task trigger time is less than V2 If the value is T, then T is updated to R. 1 The task trigger time is set; otherwise, the RGV2 car completes its transportation task, returns to idle, updates the RGV2 car status V2 to Null, and proceeds to S30142. RGV1 car simulates execution R 1 Task: Update the status of RGV1 car V1= ,in It is R 1 The task number; It is R 1 Machine tool docking station number; The value is T; The value is + ; The value is +Unloading time of raw materials; The value is + Finally, R 1 Task from R 1 Delete from the queue; S30145, if RGV2 trolley is idle and R 2 If the queue is not empty, then start from R. 2 Take a task R from the head of the queue. 2 If R 1 The queue is empty or R 2 The task trigger time is less than V1 If the value is T, then T is updated to R. 2 The task trigger time is set; otherwise, the RGV1 car completes its transportation task, returns to idle, updates the RGV1 car status V1 to Null, and proceeds to S30142. RGV2 car simulates execution R 2 Task: Update the status of RGV2 car V2= ,in It is R 2 The task number; It is R 2 Machine tool docking station number; The value is T; The value is +Time to load raw materials at the starting point of the trolley+ ; The value is + Loading time for finished products; The value is +Uninstallation time+ + Finally, R 2 Task from R 2 Delete from the queue; S30146, If neither the state V1 of RGV1 nor the state V2 of RGV2 is null, then determine whether there is a conflict between the two RGV vehicles during operation, so as to decide whether one vehicle needs to give way and wait for the other two vehicles. S30147, if neither V1 nor V2 of the vehicle is null, then compare... and ,when When the value of T is smaller, it is set to... And RGV1 car is idle and V1 is set to Null; when When the value of T is smaller, it is set to... Furthermore, RGV2 is idle and V2 is set to null; if the vehicle state V1 is not null and V2 is null, then the value of T is set to... Furthermore, if RGV1 is idle and V1 is set to null; if the vehicle state V2 is not null and V1 is null, then the value of T is set to... Furthermore, RGV2 is set to idle and V2 is set to null; the total completion time makespan is returned as T. T1.

[0027] Furthermore, in S30146, the conditions under which the two RGV vehicles conflict during operation include: spatial conflict conditions, and the state V1 of RGV1 vehicle... Less than RGV2, the car's state V2 This means that the travel paths of the two cars overlap; under the condition of temporal conflict and spatial conflict, the RGV1 car starts from the station. Arrive at the station and on the site The time interval Q1 for unloading raw materials is [ , The RGV2 car departs from the station. Arrive at the station and on the site The time interval Q2 for loading and leaving the finished product is [ , A time conflict refers to an overlap between intervals Q1 and Q2. If two RGV vehicles simultaneously meet both spatial and temporal conflict conditions, then a conflict exists between the two RGV vehicles during operation. The strategy for handling the conflict is... Less than Then the RGV2 car is numbered Station 1 is waiting to be cleared, with a waiting time w. And the RGV2 car status V2 is updated to Otherwise, the RGV1 car will be numbered. Station 1 is waiting to be cleared, with a waiting time w. And the RGV1 car status V1 is updated to .

[0028] The above-described single-rail dual-car scheduling method for the cold rolling process of seamless stainless steel tubes has at least the following beneficial effects: (1) Dynamic task selection and scenario-based scheduling mechanism. Based on machine tools , The real-time values ​​are used to filter the k highest priority tasks, ensuring that scheduling focuses on the "most urgent needs". Scheduling strategies are formulated for three scenarios based on whether R1 and R2 have tasks, avoiding inefficiency caused by a single logic, especially in the collaborative scheduling of dual task sets, and solving the parallel conflict problem of dual RGV.

[0029] (2) A dual RGV conflict detection and avoidance algorithm was established. A dual conflict determination criterion of spatial overlap and temporal overlap was proposed to accurately identify running conflicts on straight tracks. Based on the RGV departure time from the target station ( , Based on the size of the conflict, formulate differentiated site avoidance strategies, clarify the method for calculating waiting time, and ensure the orderly and quantifiable nature of conflict handling.

[0030] (3) A genetic algorithm optimization method for dual task sets was set up. A genetic algorithm was designed for dual task sets R1 and R2, with the goal of "minimizing the total completion time" to achieve global optimization of the task order. The initial population was generated independently and randomly for R1 and R2, taking into account both diversity and specificity, and avoiding premature convergence; the fitness function was directly related to the optimization objective to ensure that the algorithm evolution direction was correct.

[0031] (4) A six-tuple state and time calculation model for RGV was designed. The six-tuple fully describes the entire process of RGV from starting the task → arriving at the station → leaving the station → returning to the starting point, providing a unified data standard for scheduling simulation and conflict detection. The time calculation logic for loading and unloading tasks is distinguished, and the conflict avoidance time is superimposed to accurately quantify the time consumption of each step, providing data support for the calculation of the total completion time and scheduling optimization.

[0032] It should be understood that the exemplary embodiments described herein are illustrative and not restrictive. Although one or more embodiments of the invention have been described in conjunction with the accompanying drawings, those skilled in the art will understand that various changes in form and detail may be made without departing from the spirit and scope of the invention as defined by the appended claims.

Claims

1. A single-track dual-car scheduling method for the cold rolling process of seamless stainless steel tubes, characterized in that, Includes the following steps: S10, Set the RVG transport route: Along the transport route, set up warehouse station S1, second station S2, and multiple machine tool stations S3 to S4 in sequence. n-1 and raw material storage site S n The first trolley, RGV1, is used to transport raw materials to various machine tools, and the second trolley, RGV2, is used to transport finished products to the warehouse. The operating path of the first trolley, RGV1, is as follows: Raw material warehouse station S n As the starting point, first load the raw materials, then proceed to station S. n After unloading the raw materials at the requested machine tool station, the vehicle continues to the raw material warehouse station S at the starting point. n Stop and wait to execute the next transport task; the running path of the second trolley RGV2 is as follows: starting from the second station S2, travel to the requesting machine tool station, load the finished product, then travel to the warehouse station S1, unload the finished product, continue to travel to the second station S2 at the starting point, stop and wait to execute the next transport task; S20, triggering transportation requirements based on machine tool operating status: when the machine tool is short of raw materials, triggering a raw material transportation request; When the machine tool's finished product buffer is full, a finished product delivery request is triggered. This represents the remaining time for machine tool i to finish processing the raw material at time t. This represents the remaining time before machine tool i's finished product buffer is full at time t. and It decreases with time t; when When =0, it indicates that the machine tool is short of raw materials, triggering a raw material loading request, which is recorded as . ;when When the timer value is 0, the finished product buffer of the machine tool is full, triggering a finished product unloading request. The machine tool continues to operate. When machine tool i is replenished with raw materials at time t, then... When machine tool i removes the finished product at time t, then ; S30, perform RGV dynamic scheduling: At time t, the remaining time for each machine tool to finish processing raw materials and the remaining time for the finished product buffer to be full are respectively... , , , ... , Select the k smallest remaining time intervals. The raw material or finished product loading / unloading requests that will be triggered after the k remaining time intervals expire constitute the transportation task candidate set R. For the task candidate set R, it is divided into R based on the RGV of the task execution. 1 and R 2 Two candidate subsets of tasks are given, representing the candidate subsets of raw material transportation tasks and the candidate subsets of finished product transportation tasks, respectively, and denoted as R. 1 and R 2 The number of elements are k1 and k2 respectively, and k1 + k2 = k; R 1 and R 2 Any arrangement of tasks in the candidate set constitutes an RGV scheduling scheme, where RGV1 and RGV2 are respectively arranged according to R... 1 and R 2 The task order of the candidate set is dynamically scheduled using RGV.

2. The single-rail dual-car scheduling method for the cold rolling process of seamless stainless steel tubes as described in claim 1, characterized in that, Dynamic RGV scheduling in S30 includes: S301, if k1≠0 and k2≠0, then use a genetic algorithm to find R. 1 and R 2 Optimal task arrangement order, and based on R 1 and R 2 If the task is scheduled in the correct order, dispatch RGV1 and RGV2 vehicles to S304; otherwise, proceed to S302. S302, if k1=0 and k2≠0, then according to R 2 The tasks are sorted according to their trigger times to determine the task arrangement, and then based on R... 2 If the task is scheduled in the correct order, proceed to S304; otherwise, proceed to S303. S303, if k1≠0 and k2=0, then according to R 1 The tasks are sorted according to their trigger times to determine the task arrangement, and then based on R... 1 The task is scheduled to dispatch RGV1 trolley, then transfer to S304; S304, after scheduling m RGV vehicles, dynamically update the task candidate set and recalculate the task candidate set R.

3. The single-rail dual-car scheduling method for the cold rolling process of seamless stainless steel tubes as described in claim 2, characterized in that, In S301, a genetic algorithm is used to find R. 1 and R 2 The optimal task order includes: S3011 employs a direct permutation encoding method based on task execution order, where each chromosome represents a complete scheduling scheme, determined by R. 1 and R 2 The arrangement and composition of task numbers; S3012, the initial population is constructed using a random permutation method, for R. 1 and R 2 The task generates x distinct task permutations independently, each permutation being a random permutation of the task numbers, resulting in a population size of x. 2 ; S3013, an evaluation system is constructed based on the negative correlation between fitness and total completion time makespan. The fitness function is defined as the reciprocal of the total completion time: fitness = 1 / makespan; S3014, for a given R 1 and R 2 The task order is arranged, and the task scheduling process is simulated based on the current states of RGV1 and RGV2 vehicles to calculate the total completion time.

4. The single-track dual-car scheduling method for the cold rolling process of seamless stainless steel tubes as described in claim 3, characterized in that, The RGV status settings are as follows: The RGV state is described by a six-tuple, which is null when idle. The six-tuple includes the task ID, the task machine tool docking station ID, the current task start time, the time of arrival at the task machine tool docking station ID, the time of departure from the task machine tool docking station ID, and the time of return to the starting station of the travel path. The task ID is the task number for either the raw material transport task or the finished product transport task; the machine tool docking station ID is the machine tool docking station number that the current trolley needs to reach for the raw material transport task or the finished product transport task; the start time of the current task refers to the time when the trolley departs from the starting station of the travel path to start executing the raw material transport task or the finished product transport task, using the global system time; the time when the trolley arrives at the machine tool docking station ID is the system time when the trolley arrives at the designated machine tool docking station for the current raw material transport task or the finished product transport task; the time when the trolley leaves the machine tool docking station ID is the system time when the trolley arrives at the designated machine tool docking station, unloads raw materials or loads finished products and leaves the station; the time when the trolley returns to the starting station of the travel path is the system time when the trolley returns to the starting station after leaving the machine tool docking station, expressing the final system time when the trolley completes the current raw material transport task or the finished product transport task.

5. The single-rail dual-car scheduling method for the cold rolling process of seamless stainless steel tubes as described in claim 4, characterized in that, The process for calculating the total completion time is as follows: S30141, Obtain R 1 and R 2 The system time for triggering all tasks is determined. The current system time is obtained and denoted as T1. The initial value of the simulation scheduler clock T is set to T1. The states V1 and V2 of RGV1 and RGV2 are obtained. V1= ； V2= ； S30142, If R 1 and R 2 If all queues are empty, proceed to S30147; S30143, if RGV1 is idle and RGV2 is idle, and R 1 and R 2 If the queue is not empty, then start from R. 1 and R 2 Take one task from the head of each queue, namely R 1 With R 2 Compare R 1 With R 2 The task trigger time, if R 1 If the time is earlier, then first determine T and R. 1 The size of the task trigger time, if R 1 If the task trigger time is longer, then T is updated to R. 1 Task trigger time; Secondly, the RGV1 car simulates execution of R 1 Task: Update the status of RGV1 car V1: V1= ,in It is R 1 The task number; It is R 1 Machine tool docking station number; The value is T; The value is + ; The value is +Unloading time of raw materials; The value is + Third, when RGV2 is idle, its state V2 is set to Null; finally, R... 1 Task from R 1 Delete from the queue; If R 2 If the time is earlier, then first determine T and R. 2 The size of the task trigger time, if R 2 If the task trigger time is longer, then T is updated to R. 2 Task trigger time; secondly, the RGV2 car simulates the execution of R... 2 Task: Update the status of RGV2 car V2; V2= ,in It is R 2 The task number; It is R 2 Machine tool docking station number; The value is T; The value is +Time to load raw materials at the starting point of the trolley+ ; The value is + Loading time for finished products; The value is +Uninstallation time+ + Third, when RGV1 is idle, its state V1 is set to Null; finally, R... 2 Task from R 2 Delete from the queue; S30144, if RGV1 is idle and R 1 If the queue is not empty, then start from R. 1 Take a task R from the head of the queue. 1 If R 2 The queue is empty or R 1 The task trigger time is less than V2 If the value is T, then T is updated to R. 1 The task trigger time is set; otherwise, the RGV2 car completes its transportation task, returns to idle, updates the RGV2 car status V2 to Null, and proceeds to S30142. RGV1 car simulates execution R 1 Task: Update the status of RGV1 car V1= ,in It is R 1 The task number; It is R 1 Machine tool docking station number; The value is T; The value is + ; The value is +Unloading time of raw materials; The value is + Finally, R 1 Task from R 1 Delete from the queue; S30145, if RGV2 trolley is idle and R 2 If the queue is not empty, then start from R. 2 Take a task R from the head of the queue. 2 If R 1 The queue is empty or R 2 The task trigger time is less than V1 If the value is T, then T is updated to R. 2 The task trigger time is set; otherwise, the RGV1 car completes its transportation task, returns to idle, updates the RGV1 car status V1 to Null, and proceeds to S30142. RGV2 car simulates execution R 2 Task: Update the status of RGV2 car V2= ,in It is R 2 The task number; It is R 2 Machine tool docking station number; The value is T; The value is +Time to load raw materials at the starting point of the trolley+ ; The value is + Loading time for finished products; The value is +Uninstallation time+ + Finally, R 2 Task from R 2 Delete from the queue; S30146, If neither the state V1 of RGV1 nor the state V2 of RGV2 is null, then determine whether there is a conflict between the two RGV vehicles during operation, so as to decide whether one vehicle needs to give way and wait for the other two vehicles. S30147, if neither V1 nor V2 of the vehicle is null, then compare... and ,when When the value of T is smaller, it is set to... And RGV1 car is idle and V1 is set to Null; when When the value of T is smaller, it is set to... Furthermore, RGV2 is idle and V2 is set to null; if the vehicle state V1 is not null and V2 is null, then the value of T is set to... Furthermore, if RGV1 is idle and V1 is set to null; if the vehicle state V2 is not null and V1 is null, then the value of T is set to... Furthermore, RGV2 is set to idle and V2 is set to null; the total completion time makespan is returned as T. T1.

6. The single-track dual-car scheduling method for the cold rolling process of seamless stainless steel tubes as described in claim 5, characterized in that, In S30146, the conditions under which the two RGV vehicles may conflict during operation include: spatial conflict conditions, and the state V1 of RGV1 vehicle... Less than RGV2, the car's state V2 This means that the travel paths of the two cars overlap; under the condition of temporal conflict and spatial conflict, the RGV1 car starts from the station. Arrive at the station and on the site The time interval Q1 for unloading raw materials is [ , The RGV2 car departs from the station. Arrive at the station and on the site The time interval Q2 for loading and leaving the finished product is [ , A time conflict refers to an overlap between intervals Q1 and Q2. If two RGV vehicles simultaneously meet both spatial and temporal conflict conditions, then a conflict exists between the two RGV vehicles during operation. The strategy for handling the conflict is... Less than Then the RGV2 car is numbered Station 1 is waiting to be cleared, with a waiting time w. And the RGV2 car status V2 is updated to Otherwise, the RGV1 car will be numbered. Station 1 is waiting to be cleared, with a waiting time w. And the RGV1 car status V1 is updated to .

Images