AGV System Task Scheduling Method Based on Task Travel Time Prediction

Abstract

The present invention relates to a task scheduling method for an AGV system based on task travel time prediction, belonging to the field of intelligent manufacturing. First, when the AGV system meets the task assignment trigger condition, it is judged whether the task set to be assigned in the AGV system is empty; secondly, in the case where the task set to be assigned is not empty, according to the task scheduling scheme formulation rule, a task scheduling scheme is formulated for the task set to be assigned, and the task information in the AGV system is obtained; thirdly, the type of the AGV task is judged, and the task travel time prediction value is determined according to the travel time prediction method for different types of AGV tasks; when the task travel time prediction values of all tasks in the task set to be assigned are determined, the delay rate is calculated, and according to the task scheduling scheme evaluation rule, it is judged whether the task scheduling scheme is optimal. This method takes into account the task travel time of the AGV system, solves the problem that the task travel time in the prior art cannot be obtained and updated online and continuously, and improves the efficiency and punctuality rate of the AGV system.

Description

Technical Field

[0001] The present invention relates to the field of intelligent control, and particularly to a system task scheduling method based on task travel time prediction. Background Art

[0002] An Automatic Guided Vehicle (AGV) is a common material handling device in a workshop. According to the rated load, it can be divided into single-load AGV and multi-load AGV. A single-load AGV can only perform one loading and unloading operation in one handling task, and is suitable for transporting large or semi-finished products; a multi-load AGV can load and unload multiple trailers, and in one material task, it can load multiple loads from multiple material issuing points respectively, or unload multiple loads at multiple material delivery points, and is suitable for transporting small parts or auxiliary materials.

[0003] In an AGV material handling system, in order to ensure the completeness of materials and the punctuality of distribution, it is necessary to accurately predict the travel time of tasks. Constrained by the task travel time, the materials should be delivered before the task deadline. However, the task travel time is rich in dynamics and uncertainty and is difficult to predict. Moreover, by referring to the existing literature related to AGV material handling, there is little research on the problem of predicting the task travel time of the scheduling system, mainly focusing on the optimization problem of AGV scheduling strategies. For example, in the Chinese invention patent with the publication number CN111653098A, there is no relevant research literature on predicting the task travel time of AGV tasks. Summary of the Invention

[0004] To solve the technical defects in the prior art, the present invention proposes an AGV system task scheduling method based on task travel time prediction. By predicting the task travel time of an Automated Guided Vehicle (AGV), the problem that the task travel time in the prior art cannot be obtained and updated online and continuously is solved, and the efficiency and punctuality rate of the system are improved.

[0005] The present invention is realized through the following technical solutions:

[0006] An AGV system task scheduling method based on task travel time prediction includes the steps of:

[0007] Step 1: When the AGV system meets the task assignment trigger condition, determine whether the set of tasks to be assigned in the AGV system is empty. If it is, enter Step 1; if not, enter Step 2;

[0008] Step 2: According to the rules for formulating the task scheduling plan, formulate a task scheduling plan for the set of tasks to be assigned, and obtain the task information in the AGV system;

[0009] Step 3: Sequentially determine whether the type of the tasks in the task set to be assigned is a single-load AGV task. If it is, proceed to Step 4; if not, proceed to Step 5;

[0010] Step 4: According to the single-load AGV task travel time prediction method, determine the predicted value of the single-load AGV task travel time, and proceed to Step 6;

[0011] Step 5: According to the multi-load AGV task travel time prediction method, determine the predicted value of the multi-load AGV task travel time;

[0012] Step 6: Determine whether the predicted values of the travel times of all the tasks in the task set to be assigned have been determined. If so, proceed to Step 7; if not, proceed to Step 3;

[0013] Step 7: Based on the predicted values of the task travel times, calculate the task completion time and obtain the delay rate;

[0014] Step 8: According to the task scheduling scheme evaluation rules, determine whether the task scheduling scheme is optimal. If so, proceed to Step 1; if not, proceed to Step 2.

[0015] Furthermore, the task assignment trigger conditions include periodic trigger and event trigger.

[0016] Furthermore, the task set to be assigned is a set composed of all the tasks that have been generated and not assigned to the AGV for handling in the AGV system.

[0017] Furthermore, the task information in the AGV system includes the type, handling starting point, handling ending point, task start loading time, and task deadline of the tasks that have been assigned to the AGV and not completed handling.

[0018] Furthermore, the task scheduling scheme formulation rules include task assignment rules, path planning rules, and scheme adjustment rules.

[0019] Furthermore, the single-load AGV task travel time prediction method includes the following steps:

[0020] Step 1.1: According to the single-load AGV task free travel time calculation method, determine the predicted value of the single-load AGV task free travel time;

[0021] Step 1.2: According to the traffic delay time calculation method, determine the predicted value of the current task traffic delay time;

[0022] Step 1.3: Add the predicted value of the single-load AGV task free travel time and the predicted value of the current task traffic delay time to obtain the predicted value of the single-load AGV task travel time.

[0023] Furthermore, the calculation method for the free travel time of the single-load AGV task is as follows:

[0024]

[0025] In the formula, i, j, and k are respectively the numbers of the path segments, straight intersections, and turning intersections that the single-load AGV needs to pass through during the period from the current task loading to unloading, and I, J, and K are respectively the numbers of the path segments, straight intersections, and turning intersections that the single-load AGV needs to pass through during the period from the current task loading to unloading. t d1 and t d2 are respectively the task loading time and task unloading time of the single-load AGV, and t i , t j , and t k are respectively the free travel times of the single-load AGV passing through the i-th single path segment, the j-th straight intersection, and the k-th turning intersection.

[0026] Furthermore, the calculation methods for t i , t j , and t k are respectively as follows:

[0027]

[0028]

[0029]

[0030] In the formula, s i (i = 1, 2... I), s j (j = 1, 2... J), and s k (k = 1, 2... K) are respectively the lengths of the i-th path segment, the j-th straight intersection, and the k-th turning intersection, v i , v j , and v k are respectively the speed scalars of the AGV in the path segment, straight intersection, and turning intersection, and γ is the speed factor of the single-load AGV.

[0031] Furthermore, the calculation method for the speed factor γ of the single-load AGV is as follows:

[0032]

[0033] In the formula, F s is the current load of the single-load AGV, F es is the rated load of the single-load AGV, and α is the load coefficient of the single-load AGV.

[0034] Furthermore, α is a positive number less than 1.

[0035] Furthermore, the method for predicting the travel time of a multi-load AGV task includes the following steps:

[0036] Step 2.1: Determine the predicted value of the free travel time of the multi-load AGV task according to the calculation method of the free travel time of the multi-load AGV task.

[0037] Step 2.2: Determine the predicted value of the traffic delay time of the current task according to the calculation method of the traffic delay time.

[0038] Step 2.3: Add the predicted value of the free travel time of the multi-load AGV task and the predicted value of the traffic delay time of the current task to obtain the predicted value of the travel time of the multi-load AGV task.

[0039] Furthermore, the calculation method of the free travel time of the multi-load AGV task is as follows:

[0040]

[0041] In the formula, m, n, and p are the numbers of the path segments, straight intersections, and turning intersections that the multi-load AGV needs to pass through during the period from the current task loading to unloading, and M, N, and P are the numbers of the path segments, straight intersections, and turning intersections that the multi-load AGV needs to pass through during the period from the current task loading to unloading. T d1 、T d2 are the single-task loading time and single-task unloading time of the multi-load AGV respectively, μ1 and μ2 are the numbers of the loading tasks and unloading tasks that the multi-load AGV needs during the period from the current task loading to unloading, and T m 、T n 、T p are the free travel times of the multi-load AGV passing through the m-th path segment, the n-th straight intersection, and the p-th turning intersection respectively.

[0042] Furthermore, the calculation methods of T m 、T n 、T p are respectively:

[0043]

[0044]

[0045]

[0046] In the formula, s m (m = 1, 2...M), s n (n = 1, 2...N), s p(p = 1, 2... P) are the lengths of the m-th path segment, the n-th straight intersection, and the p-th turning intersection respectively, and δ is the speed factor of the multi-load AGV.

[0047] Furthermore, the calculation method of the speed factor δ of the multi-load AGV is as follows:

[0048]

[0049] In the formula, t and Q are the numbers and quantities of the trailers when the multi-load AGV passes through a certain path segment, straight intersection, or turning intersection respectively, and ε t is the speed factor of the t-th trailer.

[0050] Furthermore, the calculation method of the speed factor ε of the t-th trailer t is as follows:

[0051]

[0052] In the formula, F t is the current load of the t-th trailer, F et is the rated load of the t-th trailer, and η is the load coefficient of the multi-load AGV.

[0053] Furthermore, the load coefficient μ of the multi-load AGV is a positive number less than 1.

[0054] Furthermore, the calculation method of the traffic delay time is as follows:

[0055] t d = t ds + t dc (12)

[0056] In the formula, t d is the traffic delay time, t ds and t dc are the cumulative delay times of the handling AGV of the task at the straight intersection and the turning intersection respectively, and the calculation methods are as follows:

[0057]

[0058]

[0059] In the formula, s and c are the numbers of the straight intersections and turning intersections that the AGV needs to pass through during the period from the current task loading to unloading respectively, S and C are the numbers of the straight intersections and turning intersections that the AGV needs to pass through during the period from the current task loading to unloading respectively, a, Q S are the numbers and quantities of the AGVs queuing at the intersection when the AGV arrives at the intersection respectively, l a is the length of the a-th AGV queuing at the intersection, β is the safety distance coefficient, τ ais the speed factor of the a-th AGV queuing at the intersection.

[0060] Furthermore, the speed factor τ a is:

[0061]

[0062] Furthermore, the predicted value of the task travel time is calculated as follows:

[0063] T y = T fF + t d (16)

[0064] In the formula, T y is the predicted value of the task travel time, and T fF is the free travel time of the task.

[0065] Furthermore, the free travel time T of the task fF is:

[0066]

[0067] Furthermore, the task completion time is calculated as follows:

[0068] T fe = S e + T y (18)

[0069] In the formula, T fe is the task completion time, and S e is the task start loading time.

[0070] Furthermore, the calculation method of the delay rate is:

[0071]

[0072] In the formula, N is the total number of tasks in the task set to be assigned, and D e is the delay value of the e-th task.

[0073] Furthermore, the calculation method of the delay value D of the e-th task e is:

[0074]

[0075] In the formula, T e is the delay time of the e-th task, and the calculation method of the T e is:

[0076] T e = T fe - de where e = 1, …, N (21)

[0077] In the formula, d e is the task deadline of the e-th task.

[0078] Furthermore, the task scheduling scheme evaluation rule is a hybrid rule composed of the upper limit rule of the number of iterations, the rule that there is no optimization for the delay rate after multiple iterations, and the rule that the delay rate is zero. When a certain rule in the hybrid rule is satisfied, the current task scheduling scheme is optimal.

[0079] Compared with the prior art, the present invention has at least the following beneficial effects or advantages:

[0080] The present invention proposes an AGV system task scheduling method based on task travel time prediction. When performing task scheduling, the influence of task travel time is considered, solving problems such as multiple task delays and unbalanced paths that may be caused by traditional simple heuristic scheduling rules, and enabling online and continuous acquisition and update of task travel time, improving the efficiency and punctuality rate of the AGV system. BRIEF DESCRIPTION OF THE DRAWINGS

[0081] The following will further describe the present invention in detail with reference to the accompanying drawings:

[0082] Figure 1 is a flowchart of the AGV system task scheduling method based on task travel time prediction of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0083] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0084] Since the scheduling systems of single-load and multi-load automated guided vehicles are relatively complex, the following assumptions are made for the conveying system:

[0085] 1. All multi-load AGVs are centrally loaded at fixed locations, and then the multi-load AGVs perform uniform transportation according to the planned unloading sequence;

[0086] 2. Once an AGV starts transportation, it will not autonomously change the path or adjust the task unloading sequence;

[0087] 3. Idle multi-load AGVs stay at specific task waiting points waiting for the next material task, and do not affect the passage of other multi-load AGVs in the path network;

[0088] 4. The system path network is one-way single-lane and does not support overtaking.

[0089] 5. The multi-capacity AGV loads the trailers in sequence from front to back and unloads the trailers in the way of last-in-first-out.

[0090] 6. The unloading time of the multi-capacity AGV is fixed. If there are multiple AGVs at the same loading and unloading point, they need to queue up and wait.

[0091] 7. In the process of traffic control, the method of path replanning is not adopted to avoid the movement conflict of the AGV system.

[0092] Figure 1 The figure is a flowchart of the task scheduling method for the AGV system based on task travel time prediction according to the present invention. The task scheduling method for the AGV system based on task travel time prediction specifically includes the following steps:

[0093] Step 1: When the AGV system meets the task assignment trigger condition, judge whether the set of tasks to be assigned in the AGV system is empty. If it is, enter Step 1; if not, enter Step 2.

[0094] When the AGV system calls the task scheduling method for the AGV system based on task travel time prediction, it will first check whether it meets the task assignment trigger condition. When it meets the task assignment trigger condition, it judges whether the set of tasks to be assigned in the AGV system is empty. The set of tasks to be assigned is a set composed of all the tasks that have been generated in the AGV system and have not been assigned to the AGV for handling. If the set of tasks to be assigned is empty, at this time, there is no need to assign the tasks to be carried by the AGV, so there is no need to perform task scheduling, and it enters Step 1 to wait for the next time the system meets the task assignment trigger condition. If the set of tasks to be assigned is not empty, it enters Step 2.

[0095] In the embodiment of the present invention, the task assignment trigger conditions include periodic trigger and event trigger. Among them, the periodic trigger is triggered regularly at a fixed time interval, and the event trigger is to trigger task assignment when a set event occurs. In the embodiment of the present invention, the set event includes that the number of tasks in the set of tasks to be assigned reaches a certain value.

[0096] It should be noted that the task assignment trigger condition can also only select periodic trigger or event trigger, and the set event in the event trigger can also not be the set event in the embodiment of the present invention. For example, the set event can also be that there is a multi-capacity AGV task in the set of tasks to be assigned.

[0097] Step 2: According to the task scheduling scheme formulation rules, formulate a task scheduling scheme for the set of tasks to be assigned, and obtain the task information in the AGV system.

[0098] When the AGV system meets the task assignment trigger condition and the task set to be assigned is not empty, according to the task scheduling scheme formulation rules, a task scheduling scheme is formulated for the tasks in the task set to be assigned. The task scheduling scheme is to assign AGVs and path planning for all tasks in the task set to be assigned, that is, to determine the handling AGV and handling path of the tasks. After completing the formulation of the task scheduling scheme, the task information in the AGV system is obtained, including the type, handling starting point, handling ending point, task start loading time, and task deadline of the tasks that have been assigned AGVs and have not completed handling.

[0099] Among them, the task scheduling scheme formulation rules include task assignment rules, path planning rules, and scheme adjustment rules. Among them, the task assignment rules are used to assign AGVs to all tasks in the task set to be assigned, the path planning rules are used to plan paths for the tasks in the task set to be assigned, and the scheme adjustment rules are to reassign AGVs or replan paths for delayed tasks to reduce the delay rate.

[0100] Step 3: Sequentially determine whether the type of the tasks in the task set to be assigned is a single-load AGV task. If so, go to Step 4; if not, go to Step 5.

[0101] The tasks in the AGV system are divided into single-load AGV tasks and multi-load AGV tasks. Single-load AGV tasks are transported by single-load AGVs, and multi-load AGV tasks are transported by multi-load AGVs. If the task currently judged in the task set to be assigned is a single-load AGV task, go to Step 4; if not, go to Step 5.

[0102] Step 4: According to the single-load AGV task travel time prediction method, determine the predicted value of the single-load AGV task travel time, and go to Step 6.

[0103] When the task currently judged in the task set to be assigned is a single-load task, based on the task information obtained in the AGV system, the predicted value of the single-load AGV task travel time is calculated according to the single-load AGV task travel time prediction method. After obtaining the predicted value of the task travel time, go to Step 6.

[0104] Step 5: According to the multi-load AGV task travel time prediction method, determine the predicted value of the multi-load AGV task travel time.

[0105] When the task currently judged in the task set to be assigned is not a single-load task, based on the task information obtained in the AGV system, the predicted value of the multi-load AGV task travel time is calculated according to the multi-load AGV task travel time prediction method. After obtaining the predicted value of the task travel time, go to Step 6.

[0106] Step 6: Determine whether the task travel time prediction values of all tasks in the task set to be assigned are determined. If so, go to Step 7; if not, go to Step 3.

[0107] Check whether the task travel time prediction values of all tasks in the task set to be assigned have been calculated. If so, go to Step 7; if not, it means that there are still tasks in the task set to be assigned that need to calculate the task travel time prediction values. Go to Step 3 to judge the type of the next task.

[0108] Step 7: Calculate the task completion time based on the task travel time prediction value and obtain the delay rate.

[0109] After the task travel time prediction values of all tasks in the task set to be assigned are determined, calculate the task completion time of the above tasks. Based on obtaining the task completion time of the above tasks, calculate the delay rate of the current task scheduling plan. If the task completion time is later than the task deadline, then the task is a delayed task. The delay rate is the proportion of the number of delayed tasks in the task set to be assigned to the total number of tasks according to the task scheduling plan. The lower the delay rate, the better the task scheduling plan.

[0110] Step 8: According to the task scheduling plan evaluation rules, judge whether the task scheduling plan is the optimal one. If so, go to Step 1; if not, go to Step 2.

[0111] After obtaining the delay rate of the task scheduling plan, according to the task scheduling plan evaluation rules, judge whether the task scheduling plan is the optimal plan: If so, use this task scheduling plan as the final task scheduling plan for the task set to be assigned, go to Step 1, and wait for the next time the system meets the task assignment trigger condition; if not, go to Step 2, and re - formulate the task scheduling plan for the task set to be assigned according to the task scheduling plan formulation rules.

[0112] In this implementation plan, the task scheduling plan evaluation rules are a mixed rule composed of the upper limit rule of the number of iterations, the rule of no optimization of the delay rate after multiple iterations, and the rule of zero delay rate. When a certain rule in the mixed rule is met, the current task scheduling plan is the optimal one. The upper limit rule of the number of iterations means that when the total number of task scheduling plans formulated for the current task set to be assigned reaches a certain value, the last task scheduling plan is the optimal plan. The rule of no optimization of the delay rate after multiple iterations means that when the delay rates of the task scheduling plans of the current task set to be assigned remain unchanged for a certain number of consecutive times, the last task scheduling plan is the optimal plan. The rule of zero delay rate means that if the delay rate of the task scheduling plan is 0, then the task scheduling plan is the optimal plan.

[0113] It should be noted that the task scheduling plan evaluation rules can also adopt one or more of the above rules, or can also adopt rules other than the above.

[0114] Single-load AGV task travel time prediction method, including the following steps:

[0115] Step 1.1: Determine the predicted value of the free travel time of the single-load AGV task according to the free travel time calculation method of the single-load AGV task;

[0116] The free travel time of the single-load AGV task is jointly affected by various factors such as the AGV movement speed, running path, loading and unloading time, AGV kinematics and dynamics characteristics, running path segment characteristics, load state, etc. In the embodiments of the present invention, the AGV kinematics and dynamics characteristics include the AGV straight-line speed, turning speed, acceleration and deceleration characteristics, etc., the running path segment characteristics include path segments, straight intersections, turning intersections, and the load state includes no load, full load, loaded, etc. The free travel time of the single-load AGV task is specifically obtained by the free travel time calculation method of the single-load AGV task.

[0117] The free travel time calculation method of the single-load AGV task is:

[0118]

[0119] In the formula, i, j, and k are the numbers of the path segment, straight intersection, and turning intersection that the single-load AGV needs to pass through during the period from the current task loading to unloading, respectively. I, J, and K are the numbers of the path segment, straight intersection, and turning intersection that the single-load AGV needs to pass through during the period from the current task loading to unloading, respectively. t d1 、t d2 are the task loading time and task unloading time of the single-load AGV, respectively. t i 、t j 、t k are the free travel times of the single-load AGV passing through the i-th single path segment, the j-th straight intersection, and the k-th turning intersection, respectively. t i 、t j 、t k The calculation methods are as follows:

[0120]

[0121]

[0122]

[0123] In the formula, s i (i = 1, 2...I), s j (j = 1, 2...J), s k (k = 1, 2...K) are the lengths of the i-th path segment, the j-th straight intersection, and the k-th turning intersection, respectively. v i 、v j 、vk They are the speed scalars of the AGV at the path segment, straight intersection, and turning intersection respectively. γ is the speed factor of the single-load AGV, and its calculation method is as follows:

[0124]

[0125] In the formula, f s is the current load of the single-load AGV, F es is the rated load of the single-load AGV, and α is the load coefficient of the single-load AGV. α is a positive number less than 1.

[0126] In the embodiment of the present invention, according to experience, the value range of α is 0.8 - 0.9.

[0127] Step 1.2: Determine the predicted value of the current task traffic delay time according to the traffic delay time calculation method;

[0128] The traffic delay time is affected by factors such as the length of the AGV itself, the safety distance between AGVs, and the length of a single path, and is specifically obtained by the traffic delay time calculation method. The traffic delay time calculation method is applicable to both single-load AGV tasks and multi-load AGV tasks, and the traffic delay time calculation method will be further described later.

[0129] Step 1.3: Determine the predicted value of the task travel time according to the task travel time prediction value calculation method.

[0130] The predicted value of the task travel time can be obtained by the task travel time prediction value calculation method and is jointly determined by the free travel time of the task and the traffic delay time. The free travel time of the task refers to the free running time of the AGV in the path network without interference from other AGVs and is the basis of the task travel time. The traffic delay time refers to the waiting time of the AGV at the intersection when there are many AGVs in the system and congestion occurs. Considering the free travel time of the task and the traffic delay time ensures the accuracy of the predicted task travel time. The task travel time prediction value calculation method is applicable to both single-load AGV tasks and multi-load AGV tasks, and the task travel time prediction value calculation method will be further described later.

[0131] The method for predicting the travel time of a multi-load AGV task includes the following steps:

[0132] Step 2.1: Determine the predicted value of the free travel time of the multi-load AGV task according to the free travel time calculation method of the multi-load AGV task;

[0133] The factors affecting the free travel time of multi-load AGV tasks are the same as those of single-load AGV tasks. However, in terms of the running path, the running path of a single-load AGV depends on the loading and unloading points of the task, while the running path of a multi-load AGV not only depends on the loading and unloading points of all trailers but is also closely related to the loading and unloading sequence of the trailers.

[0134] In the embodiments of the present invention, the calculation method for the free travel time of multi-load AGV tasks is as follows:

[0135]

[0136] In the formula, m, n, and p are the numbers of the path segments, straight intersections, and turning intersections that the multi-load AGV needs to pass through during the loading to unloading of the current task, respectively; M, N, and P are the numbers of the path segments, straight intersections, and turning intersections that the multi-load AGV needs to pass through during the loading to unloading of the current task, respectively; T d1 and T d2 are the single-task loading time and single-task unloading time of the multi-load AGV, respectively; μ1 and μ2 are the numbers of loading tasks and unloading tasks that the multi-load AGV needs to perform during the loading to unloading of the current task, respectively; T m and T n and T p are the free travel times of the multi-load AGV for the tasks passing through the m-th path segment, the n-th straight intersection, and the p-th turning intersection, respectively; T m and T n and T p The calculation methods are as follows:

[0137]

[0138]

[0139]

[0140] In the formula, s m (m = 1, 2...M), s n (n = 1, 2...N), s p (p = 1, 2...P) are the lengths of the m-th path segment, the n-th straight intersection, and the p-th turning intersection, respectively; δ is the speed factor of the multi-load AGV, and its calculation method is:

[0141]

[0142] In the formula, t and Q are the numbers of the trailers when the multi-load AGV passes through a certain path segment, straight intersection, or turning intersection, respectively; ε t is the speed factor of the t-th trailer, and its calculation method is:

[0143]

[0144] In the formula, F t is the current load of the t-th trailer, and F et is the rated load of the t-th trailer. η is the load coefficient of the overloaded AGV, and μ is a positive number less than 1.

[0145] In the embodiment of the present invention, according to experience, the value range of μ is 0.8 - 0.9.

[0146] Step 2.2: Determine the predicted value of the current task traffic delay time according to the traffic delay time calculation method;

[0147] The traffic delay time calculation method is as follows:

[0148] t d = t ds + t dc (12)

[0149] In the formula, t d is the traffic delay time, and t ds , t dc are respectively the cumulative delay times of the handling AGV of the task at the straight intersection and the turning intersection. The calculation methods are as follows:

[0150]

[0151]

[0152] In the formula, s and c are respectively the numbers of the straight intersections and turning intersections that the AGV needs to pass through during the period from loading to unloading of the current task, S and C are respectively the numbers of the straight intersections and turning intersections that the AGV needs to pass through during the period from loading to unloading of the current task, a, Q S are respectively the number and quantity of the AGVs queuing at the intersection when the AGV arrives at the intersection, l a is the length of the a-th AGV queuing at the intersection, β is the safety distance coefficient, and τ a is the speed factor of the a-th AGV queuing at the intersection, and its specific value is:

[0153]

[0154] Step 2.3: Determine the predicted value of the task travel time according to the task travel time prediction value calculation method.

[0155] The task travel time prediction value calculation method is as follows:

[0156] T y = T fF + t d (16)

[0157] In the formula, T y is the predicted value of the task travel time, and T fF is the task free travel time.

[0158] The task free travel time T fF is:

[0159]

[0160] In the embodiment of the present invention, the task completion time is calculated as follows:

[0161] T fe = S e + T y (18)

[0162] In the formula, T fe is the task completion time, and S e is the task start loading time.

[0163] It should be noted that the calculation method of the task completion time is applicable to both single-load AGV tasks and multi-load AGV tasks.

[0164] In the embodiment of the present invention, the delay rate is calculated as follows:

[0165]

[0166] In the formula, N is the total number of tasks in the task set to be allocated, and D e is the delay value of the e-th task, and its calculation method is:

[0167]

[0168] In the formula, T e is the delay time of the e-th task, and its calculation method is:

[0169] T e = T fe - d e , e = 1,..., N (21)

[0170] In the formula, d e is the task deadline of the e-th task.

[0171] The above are only the preferred embodiments of the present invention. It should be noted that for those of ordinary skill in the art, several improvements can be made without departing from the principle of the present invention, and these improvements should also be regarded as the protection scope of the present invention.

Claims

1. A task scheduling method for an AGV system based on task travel time prediction, characterized in that, Including the steps: Step 1: When the AGV system meets the task assignment trigger condition, determine whether the task set to be assigned in the AGV system is empty. If it is, enter Step 1; if not, enter Step 2; Step 2: According to the task scheduling plan formulation rules, formulate a task scheduling plan for the task set to be assigned, and obtain the task information in the AGV system; Step 3: Sequentially determine whether the type of the task in the task set to be assigned is a single-load AGV task. If it is, enter Step 4; if not, enter Step 5; Step 4: According to the single-load AGV task travel time prediction method, determine the predicted value of the single-load AGV task travel time, and enter Step 6; Step 5: According to the multi-load AGV task travel time prediction method, determine the predicted value of the multi-load AGV task travel time; Step 6: Determine whether the predicted values of the travel times of all tasks in the task set to be assigned have been determined. If so, enter Step 7; if not, enter Step 3; Step 7: Based on the predicted values of the task travel times, calculate the task completion time and obtain the delay rate; Step 8: According to the task scheduling plan evaluation rules, determine whether the task scheduling plan is optimal. If so, enter Step 1; if not, enter Step 2.

2. The task scheduling method of the AGV system based on task travel time prediction according to claim 1, characterized in that, The task scheduling plan formulation rules include task assignment rules, path planning rules, and plan adjustment rules.

3. The task scheduling method of the AGV system based on task travel time prediction according to claim 1, wherein The task scheduling plan evaluation rules are a mixed rule composed of the upper limit rule of the number of iterations, the rule that there is no optimization for multiple iterations of the delay rate, and the rule that the delay rate is zero.

4. The task scheduling method of the AGV system based on task travel time prediction according to claim 1, characterized in that, The single-load AGV task travel time prediction method includes the following steps: Step 1.1: According to the single-load AGV task free travel time calculation method, determine the predicted value of the single-load AGV task free travel time; Step 1.2: According to the traffic delay time calculation method, determine the predicted value of the current task traffic delay time; Step 1.3: According to the task travel time prediction value calculation method, determine the predicted value of the task travel time.

5. The task scheduling method of the AGV system based on task travel time prediction according to claim 4, wherein, The single-load AGV task free travel time calculation method is: Wherein, i, j, and k are respectively the numbers of the path segments, straight intersections, and turning intersections that the single-load AGV needs to pass through during loading to unloading of the current task, and I, J, and K are respectively the numbers of path segments, straight intersections, and turning intersections that the single-load AGV needs to pass through during loading to unloading of the current task, t d1 , t d2 are respectively the task loading time and task unloading time of the single-load AGV, and t i , t j , t k are respectively the free travel times of the single-load AGV passing through the i-th single path segment, the j-th straight intersection, and the k-th turning intersection; The said t i 、t j 、t k are calculated respectively as follows: where s i (i = 1, 2… I), s j (j = 1, 2… J), s k (k = 1, 2… K) are the lengths of the i-th path segment, the j-th straight intersection, and the k-th turning intersection respectively, v i , v j , v k are the speed scalars of the AGV at the path segment, the straight intersection, and the turning intersection respectively, and γ is the speed factor of the single-load AGV; The calculation method of the single-load AGV speed factor γ is: Where, F s is the current load of the single-load AGV, F es is the rated load of the single-load AGV, and α is the load factor of the single-load AGV.

6. The task scheduling method of the AGV system based on task travel time prediction according to claim 1, characterized in that The multi-load AGV task travel time prediction method includes the following steps: Step 2.1: According to the multi-load AGV task free travel time calculation method, determine the predicted value of the multi-load AGV task free travel time; Step 2.2: According to the traffic delay time calculation method, determine the predicted value of the current task traffic delay time; Step 2.3: According to the task travel time prediction value calculation method, determine the predicted value of the task travel time.

7. The task scheduling method of the AGV system based on task travel time prediction according to claim 6, wherein, The calculation method of the multi-load AGV task free travel time is: Wherein, m, n, and p are the numbers of path segments, straight intersections, and turning intersections that the multi-load AGV needs to pass through during the period from the current task loading to unloading, M, N, and P are the numbers of path segments, straight intersections, and turning intersections that the multi-load AGV needs to pass through during the period from the current task loading to unloading, and T d1 , T d2 are the single-task loading time and single-task unloading time of the multi-load AGV respectively, μ1 and μ2 are the numbers of loading tasks and unloading tasks that the multi-load AGV needs during the period from the current task loading to unloading, T m , T n , T p are the task free travel times of the multi-load AGV passing through the m-th path segment, the n-th straight intersection, and the p-th turning intersection respectively; The said T m , T n , T p The calculation methods are respectively as follows: where s m (m = 1, 2…M), s n (n = 1, 2…N), s p (p = 1, 2…P) are the lengths of the m-th path segment, the n-th straight intersection, and the p-th turning intersection respectively, v i , v j , v k are the speed scalars of the AGV at the path segment, the straight intersection, and the turning intersection respectively, and δ is the speed factor of the multi-load AGV; The calculation method of the multi-load AGV speed factor δ is: Wherein, t and Q are respectively the number and quantity of trailers when the multi-load AGV passes through a certain path segment, a straight intersection or a turning intersection, and ε t is the speed factor of the t-th trailer; The speed factor ε of the t-th trailer t is calculated as follows: Where, F t is the current load of the t-th trailer, and F et is the rated load of the t-th trailer, and η is the load coefficient of the overloaded AGV.

8. The task scheduling method of the AGV system based on task travel time prediction according to claim 4, characterized in that, The traffic delay time calculation method is: t d = t ds + t dc (12) where t d is the traffic delay time, and t ds , t dc are the cumulative delay times of the handling AGV of the task at the straight intersection and the turning intersection respectively. The calculation methods are as follows: Wherein, s and c are respectively the numbers of straight intersections and turning intersections that the AGV needs to pass through during the period from loading to unloading of the current task, S and C are respectively the numbers of straight intersections and turning intersections that the AGV needs to pass through during the period from loading to unloading of the current task, a, Q S are respectively the numbers and quantities of AGVs queuing at the intersection when the AGV arrives at the intersection, l a is the length of the a-th AGV queuing at the intersection, β is the safety distance coefficient, τ a is the speed factor of the a-th AGV queuing at the intersection, s j (j = 1, 2... J), s k (k = 1, 2... K) are respectively the lengths of the j-th straight intersection and the k-th turning intersection, v j 、v k are respectively the speed scalars of the AGV at the straight intersection and the turning intersection; The speed factor τ a is as follows: The task travel time prediction value calculation method is: T y = T fF + t d (16) where T y is the predicted value of the task travel time, and T fF is the free travel time of the task; Further, the task free travel time T fF is as follows:

9. The task scheduling method for an AGV system based on task travel time prediction according to claim 1, wherein The calculation method of the task completion time is: T fe = S e + T y (18) In the formula, T fe is the task completion time, S e is the task start loading time, and T y is the predicted value of the task travel time.

10. The task scheduling method of the AGV system based on task travel time prediction according to claim 9, wherein, The calculation method of the delay rate is: where N is the total number of tasks in the task set to be allocated, and D e is the delay value of the e-th task; The delay value D of the e-th task e is calculated as follows: where T e is the latency of the e-th task, and the calculation method of the said T e is as follows: T e = T fe - d e , e = 1, …, N (21) where d e is the deadline of the e-th task.

Images