A Scheduling Method for AGVs in Automated Terminals

Abstract

The present invention provides a scheduling method for AGVs in an automated terminal, comprising the following steps: extracting relevant information to obtain the path information for the transportation scheduling of the AGVs in the automated terminal; establishing the objective function and constraint conditions of the AGV scheduling method in the automated terminal; establishing the overall AGV scheduling objectives and principles; generating the sparrow optimization algorithm; improving the identity transformation rule and position transformation rule of the sparrow optimization algorithm; performing iterative operations until the current iteration number reaches the set maximum iteration number; and outputting the optimal AGV scheduling scheme and the optimal value of the corresponding objective function. The AGV scheduling method provided by the present invention extracts the delivery time of the terminal goods, the node information of the shelves, and the vehicle operation information, etc., comprehensively considers the factor of the lowest empty load rate, establishes a problem model regarding cost, time, and empty load rate, and considers factors such as penalty cost to establish a reasonable scheduling principle for multiple AGVs.

Description

Technical Field

[0001] The present invention relates to a scheduling method for AGVs in an automated terminal, belonging to the technical field of logistics transportation planning. Background Art

[0002] In the process of warehousing logistics, the time consumed in links such as loading, unloading, and transportation accounts for the vast majority of the total logistics process time, resulting in a relatively high proportion of material transportation costs in the total cost. The multi-AGV path planning method plays an important role in improving the structure of the logistics transportation system, reducing the logistics transportation cost, and enhancing the system operation efficiency. Therefore, it has great application value in warehousing logistics. The path planning of AGVs mainly includes three aspects of problems: 1) determining whether there is a feasible path between the starting point and the target point; 2) the planned path must be unblocked, conflict-free, and deadlock-free; 3) the planned path should make the operation efficiency of the entire system reach the optimum.

[0003] With the rapid development of the global economy and science and technology, terminals are also developing towards automation. In a container automated terminal, the production process control system in the central control room monitors the operating status of AGVs using a wireless local area network, makes data judgments and transmissions, and issues correct task instructions. Due to the differences in performance parameters such as the mechanical equipment configuration, tolerance parameters, and utilization rate of each terminal, the AGV scheduling method needs to be adjusted accordingly. The scientific nature of the scheduling decision directly affects the carrying rate, energy consumption rate, and timeliness of AGVs. At the same time, the purpose of AGVs transporting containers is not only to carry them to the designated storage location, but also to take into account the time constraints, transportation paths, and loading costs. Therefore, the present invention starts from considering the lowest cost, shortest time, and lowest empty load rate, and reasonably plans to achieve the reasonable transportation of AGVs within the specified time window. Summary of the Invention

[0004] The purpose of the present invention is to provide a scheduling method for AGVs in an automated terminal, comprehensively considering various constraints affecting the scheduling result such as the number of AGVs, AGV path planning, multi-AGVs, and reasonable scheduling principles, and establishing an optimization model, so as to achieve more reasonable scheduling of AGVs in the automated terminal.

[0005] The technical solution of the present invention is as follows: The present invention discloses a scheduling method for AGVs in an automated terminal, including the following steps:

[0006] S1: Extract relevant information to obtain the path information for the transportation scheduling of AGVs in the automated terminal;

[0007] S2: Establish the objective function and constraint conditions of the AGV scheduling method in the automated terminal according to the extracted node information;

[0008] S3: Establish the overall AGV scheduling objectives and principles;

[0009] S4: Use the sparrow optimization algorithm to solve the AGV scheduling problem, generate the initial population of the sparrow optimization algorithm, and calculate the population fitness;

[0010] S5: Improve the identity transformation rule and position transformation rule of the sparrow optimization algorithm, and perform iterative operations until the current iteration number reaches the set maximum iteration number;

[0011] S6: Output the best AGV scheduling scheme and the optimal value of the corresponding objective function.

[0012] Furthermore, for the above scheduling method of the AGV in the automated terminal, where: The relevant information in the step S1 includes:

[0013] A{a1,a2,a3...a n} represents the set of AGVs,

[0014] H{h1,h2...h n} represents the set of shelf nodes,

[0015] The congestion waiting time of the AGV K from shelf i to shelf j is

[0016] represents the AGV K transporting from shelf i to shelf j,

[0017] represents the transportation time from shelf i to shelf j,

[0018] represents the manual operation time at shelf i,

[0019] (v i ,m i ) represents the time window for the goods to dock at the shore,

[0020] t g represents the time required for the trolley to charge during the transportation completion,

[0021] t s represents the time generated due to rescheduling the goods caused by accidents,

[0022] represents the AGV K transportation cost generated during the transportation from i to j,

[0023] represents the cost generated by manual operation at transportation point i,

[0024] c s Represents the penalty cost incurred when the transportation is not completed within the specified time.

[0025] Furthermore, for the above scheduling method of the AGV in the automated terminal, where: In step S2, the objective function includes:

[0026] Objective function 1: This objective function is based on the requirement that when using AGVs for arrival storage, the time required to install all the goods should be the shortest;

[0027] Objective function 2: This objective function is based on the lowest cost of completing the tasks, where: Represents the weighting coefficient,

[0028]

[0029] In the formula, m q is the quantity of goods, and w q is the value of the goods attribute;

[0030] Objective function 3: This objective function is based on arranging vehicles for multi-task simultaneous transportation with the lowest vehicle idle rate under the condition of meeting the time window limit and its own capacity limit, where: C is the capacity of an AGV cart, and c u , u = 1, 2, 3... n, is the space capacity occupied by each task.

[0031] Furthermore, for the above scheduling method of the AGV in the automated terminal, where: The constraint conditions in step S2 include:

[0032]

[0033] c u ≤ C (5)

[0034] Among them: Equation (1) means that the final time to complete all handling tasks does not exceed the maximum ship docking time; Equation (2) means whether the AGV cart is transporting on paths i and j; Equation (3) means that one task needs to be completed in one transportation; Equation (4) means that one container loading and unloading is completed by one AGV cart; Equation (5) represents the capacity constraint of the cart.

[0035] Furthermore, for the above scheduling method of the AGV in the automated terminal, where: In step S3, the AGV scheduling target is t f ≤ m i ; t f is the time to complete the handling,

[0036] The AGV scheduling principle is as follows:

[0037] minc1 = mod(q l,i , q r,i ); (1)

[0038] minc2 = minc s ; (2)

[0039]

[0040] lx s = lx d = l; if ▽ c (4)

[0041]

[0042] Where the total freight volume to be transported is Q l , the single-piece load of different types of goods is q r,i = {q r,1 , q r ,2 ,..., q r,N}, the shipping volume of each trolley with different loading capacities is q li = {q l,1 , q l,2 ,..., q l,N}, the vehicles moving in the transportation direction are denoted as x s , the empty trolleys returning are denoted as x d , then their priorities are lx s , lx d The time window is (v i , m i ), the number of loaded trolleys The total number of operating trolleys is m w , the situation of conflicts is denoted as ▽ c , t p represents the time required to complete a single cargo transportation;

[0043] The expressions in the formula are as follows in turn:

[0044] (1) When retrieving idle trolleys from the ship docking point, according to the required carrying capacity q1 of the goods, give priority to arranging the transportation of vehicles with a transportation capacity of q1·n, (n = 1, 2, 3...) that is, the AGV transportation capacity is a multiple of it, give priority to retrieving idle trolleys, and then retrieve charging trolleys;

[0045] (2) Restricted by the time window (v i , m i ) of ship scheduling, consider the penalty cost c s generated by not completing the transportation within the specified time to be the minimum;

[0046] (3) When planning a path in the grid network between the starting point and the ending point, there will be a trolley k1 for load transportation from the starting point to the ending point, and a trolley k2 waiting to be reloaded and returning from the ending point to the starting point. The vehicle transporting in the transportation direction is denoted as x s , and the empty return trolley is denoted as x d , when the number of trolleys with load tasks reaches half of the total number of operating trolleys, the priority of k1 is always greater than that of k2, and vice versa;

[0047] (4) When there is a conflict in trolley transportation, according to the established priority function l, reasonable vehicle scheduling is carried out:

[0048] (5) When there is a task arrangement, by arranging the number of trolleys Q A , the operating efficiency of the system is made better.

[0049] Furthermore, for the above scheduling method of the AGV in the automated terminal, where: the steps of the priority setting are as follows. Two indicators are selected. The first indicator is the remaining order quantity q of the trolley j , and the second indicator is the distance s of the shelf number j . The orders of the two indicators are l1 and l2 respectively. When congestion occurs, the following rules are followed: referring to the number m of trolleys in operation, the final priority rule l for trolley scheduling is

[0050]

[0051] where M is the number of all operable trolleys,

[0052] The meaning of this function is:

[0053] When the number of operating vehicles is less than , according to l1, that is, the trolley with more remaining order quantity has priority;

[0054] When the number of operating vehicles is more than , according to l2, that is, the trolley farther from the shelf has priority;

[0055] When the number of operating vehicles is between and , according to the above rules, the priority scheduling mainly focuses on the transportation volume.

[0056] Furthermore, for the above scheduling method of the AGV in the automated terminal, where: in step S4, an initial population of the sparrow optimization algorithm is generated

[0057] Its corresponding fitness

[0058] Furthermore, in the above scheduling method of the AGV in the automated terminal, where: in step S5, the identity transformation rule and position transformation rule of the improved sparrow optimization algorithm are as follows:

[0059] S5.1 Update the position of the discoverer:

[0060]

[0061] where: t represents the current iteration number, j = 1, 2, 3... d. R2 ∈ [0, 1] and ST ∈ [0.5, 1] represent the warning value and safety value respectively, Q is a random number conforming to the normal distribution, L represents a 1×d matrix, where each element in the matrix is all 1, R2 < ST represents less than the warning range, and R2 ≥ ST represents reaching the warning range;

[0062] S5.2 Update the position of the joiner:

[0063]

[0064] where, X p represents the optimal position occupied by the current discoverer, X worst represents the current globally worst position, A is a 1xd matrix with all elements being 1 or -1, and A + = A T (AA T ) -1 ;

[0065] S5.3 Update the position of the scout:

[0066]

[0067] where: X best is the current globally optimal position, β is the step size control parameter, which is a random number obeying the normal distribution with a mean of 0 and a variance of 1; f g and f w represent the current globally best and worst fitness respectively, K is a random number within [-1, 1], and ε is a constant to avoid the denominator being 0;

[0068] S5.4 Add the identity transformation rule for the joiner to discoverer:

[0069] Set the time t. When there is no better position under the set time t, a certain number m of joiners are directly changed to discoverers, and both t and m are dynamically changing;

[0070] Assume that the number of algorithm iterations is N and the current iteration number is n;

[0071] When it reaches the first At this time, t follows At this time, m follows log n / N+1 n;

[0072] When it comes to the end At this time, t follows At this time, m follows log N-n / N+1 n;

[0073] Among them, at the beginning When t decreases and m increases, but the growth rate of m decreases later;

[0074] S5.5 Add the position transformation rule of the joiner and discoverer:

[0075] When the discoverer discovers food, there will be multiple joiners jumping to its position at the same time, which is not conducive to expanding the search efficiency in the initial stage. Set the flight speed of each joiner, so that when one reaches the joiner's position, other sparrows are not allowed to fly over;

[0076] S5.6 Propose the relevant rules of the sparrow optimization algorithm regarding the visibility degree:

[0077] When the joiner changes its identity to the discoverer a certain number of times τ, the eyesight will decline, and the discoverer will not become invisible. The advantage is that the visualization degree is high in the early stage, expanding the search range, and the search range is reduced when the eyesight declines in the later stage, so as to improve the convergence speed of the algorithm;

[0078] S5.7 Judge whether the current iteration number n reaches the set maximum iteration number iter max , if n > iter max , then output the value f of the best fitness at this time g As the optimal target value, the corresponding one is the optimal target solution, so as to find the best AGV scheduling scheme and the corresponding minimum value Z min ; Otherwise, continue with step S5.

[0079] Furthermore, for the above scheduling method of the AGV in the automated terminal, where: The specific steps of the above step S5.5 are as follows:

[0080] Step 1: Set the flight speeds of the joiners as v1, v2... v n , and the distances from the discoverer who discovers food are d1, d2... d n ;

[0081] Step 2: When the discoverer discovers food, compare the speeds v of each joiner k ;

[0082] Step 3: If the speed v of one of the joiners k , makes d k / v k If it is less than other values, then all other joiners maintain the original operation, and this speed is v k The joiner jumps to the discoverer's position.

[0083] Furthermore, for the above-mentioned scheduling method of the automated terminal AGV, wherein: the specific steps of the step S5.6 are as follows:

[0084] Step 1: Set the vision ε of the joiner sparrow and the visibility γ of the discoverer. The two are in a positive correlation, that is, ε = kγ (k>0), The number of algorithm iterations is N, and the current iteration number is n;

[0085] Step 2: Determine whether the number of times σ that the joiner changes its identity to a discoverer reaches a certain number τ;

[0086] Step 3: If σ < τ, maintain the original search rule; if σ ≥ τ, ε changes accordingly at this time, the sparrow's vision decreases, and the corresponding discoverer's visibility also decreases.

[0087] Compared with the prior art, the present invention adopts the above technical solutions and has the following technical effects: The automated terminal AGV scheduling method provided by the present invention makes up for the deficiencies of the current AGV scheduling scheme, extracts the delivery time of terminal goods, the node information of the shelves, and the vehicle operation information, etc., comprehensively considers the factor of the lowest empty load rate, establishes a problem model regarding cost, time, and empty load rate, considers factors such as penalty cost, and establishes a reasonable scheduling principle for multiple AGVs, which has important practical significance in practical problems; Secondly, this article improves the sparrow optimization algorithm and applies it to the problem. The improvements include the position transformation rule and identity transformation rule of the joiner and the discoverer, and the consideration of constructing the sparrow's vision and visibility, so that the algorithm reduces the situation of converging to the local optimal solution, can not only quickly find the optimal solution, but also reasonably give the AGV path plan. BRIEF DESCRIPTION OF THE DRAWINGS

[0088] Figure 1 is the flow chart of the steps of the automated terminal AGV scheduling method of the present invention; DETAILED DESCRIPTION OF THE EMBODIMENTS

[0089] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only for illustration and are not used to limit the present invention.

[0090] The present invention proposes an automated terminal AGV scheduling method, as Figure 1 shown, including the following steps:

[0091] S1: Extract the delivery time of the terminal goods, the node information of the shelves, and the vehicle operation information, etc., to obtain the path information for the transportation scheduling of the AGV in this automated terminal.

[0092] S2: Establish the objective function and constraints of the AGV scheduling method in the automated terminal according to the extracted node information.

[0093] S3: According to the vehicle scheduling priority rule, reduce the proportion of the free space of each transport vehicle in the total space each time, so as to reduce the empty load rate; according to the dynamic setting method of the number of AGVs, set the number of AGV cars, thereby establishing the overall AGV scheduling principle.

[0094] S4: Use the sparrow optimization algorithm to solve the AGV scheduling problem, generate the initial population of the sparrow optimization algorithm and calculate the population fitness.

[0095] S5: Improve the identity transformation rule and position transformation rule of the sparrow optimization algorithm, and perform iterative operations until the current number of iterations reaches the set maximum number of iterations.

[0096] S6: Output the best AGV scheduling plan and the optimal value of the corresponding objective function.

[0097] Specifically, the relevant parameters involved in step S1 are as follows;

[0098] A{a1,a2,a3...a n} represents the set of AGVs,

[0099] H{h1,h2...h n} represents the set of shelf nodes,

[0100] AGV K The congestion waiting time of the shelf from i to j is

[0101] Indicates that the AGV K Transports from shelf i to shelf j,

[0102] Represents the time for transporting from shelf i to shelf j,

[0103] Represents the manual operation time at shelf i,

[0104] (v i ,m i ) represents the time window for the goods to dock at the shore,

[0105] t g Represents the time required for the trolley to charge during the completion of transportation,

[0106] t s represents the time generated by rescheduling goods due to accidents

[0107] represents the AGV K transportation cost generated during the transportation from i to j

[0108] represents the cost generated by manual operation at transportation point i

[0109] c s represents the penalty cost generated for not completing the transportation within the specified time

[0110] is a weighted weight set to give one objective priority over another

[0111] In step S2, to meet the requirements of the lowest cost, the shortest transportation time, and the lowest vehicle idle rate, the number of AGVs is dynamically set and the AGV scheduling principle is established. The objective function and constraints are as follows

[0112] Objective function 1

[0113] where: A{a1, a2, a3... a n} represents the set of AGVs, H{h1, h2... h n} represents the set of shelf nodes represents the AGV K congestion waiting time from shelf i to j represents the AGV K transporting from shelf i to shelf j represents the time for transporting from shelf i to shelf j, t g represents the time required for the trolley to charge after completing the transportation, t s represents the time generated by rescheduling goods due to accidents. This objective function is based on the requirement that when using AGVs for arrival storage, the time required to install all goods should be the shortest

[0114] Objective function 2

[0115] where: A{a1, a2, a3... a n} represents the set of AGVs, H{h1, h2... h n} represents the set of shelf nodes represents the AGV K transporting from shelf i to shelf j represents the AGV K transportation cost generated during the transportation from i to j represents the cost of manual operation at transport point i, c s Indicates the penalty cost for not completing the transportation within the specified time. The weighting of one objective over another is set, and this objective function is based on the minimum cost of completing the task.

[0116] Weighting coefficient The relationship is:

[0117]

[0118] Where m q is the quantity of goods, w q is the attribute value of the goods,

[0119] Weighting coefficient Indicates its relationship with the quantity of goods m q and the value of the goods attribute w q It is related, and it changes dynamically with the value and quantity of goods. As the value increases, the priority ratio increases, and as the quantity increases, the priority ratio decreases, mainly considering the value priority rule.

[0120] Objective function 3: When arranging AGV for transportation, the capacity of an AGV is C. If the space occupied by a transportation task is less than C, there will be surplus space, which will increase the transportation cost in disguise. The space capacity occupied by each task is c. u , u=1,2,3...n, this objective function requires that vehicles be arranged to perform multiple tasks and transport at the lowest empty rate while meeting the time window constraints and their own capacity constraints.

[0121] The constraints in step S2 include:

[0122]

[0123]

[0124] c u ≤C (5)

[0125] Among them: Formula 1 means that the final time to complete all handling tasks does not exceed the maximum stop time; Formula 2 indicates whether the AGV car is transported on paths i and j; Formula 3 means that one task requires one transportation to complete; Formula 4 means that one container loading and unloading is completed by one AGV car; Formula 5 represents the capacity constraint of the car.

[0126] The method for dynamically setting the number of AGVs in step S3 is as follows:

[0127]

[0128] Represents the time for transporting from shelf i to shelf j.

[0129] Represents the manual operation time at shelf i.

[0130] t g Represents the time required for the trolley to charge during the completion of transportation.

[0131] t s Represents the time generated due to rescheduling goods caused by accidents.

[0132] t p Represents the time required to complete a single goods transportation.

[0133] In step S3, the following method is adopted to establish the vehicle scheduling priority rule:

[0134] When the AGV trolley is carrying goods, congestion may occur and the signal may be interfered. At this time, it is necessary to establish the priority of trolley scheduling. Two indicators are selected. The first indicator is the remaining order quantity q of the trolley j , and the second indicator is the distance s of the shelf number j . The orders of the two indicators are l1 and l2 respectively. When congestion occurs, the following rules are followed: referring to the number m of running trolleys, the final priority rule l of trolley scheduling is (where Q A is the number of all operable trolleys,

[0135]

[0136] The meaning of this function is:

[0137] When the number of running vehicles is less than , according to l1, that is, the one with more remaining order quantity is preferred;

[0138] When the number of running vehicles is more than , according to l2, that is, the one far from the shelf is preferred;

[0139] When the running quantity is between and , according to the above rules, it mainly focuses on preferential scheduling based on the transportation volume.

[0140] In step S3, the scheduling principles of the AGV are proposed, as follows: When scheduling the AGV, a series of principles such as the task attributes, the urgency of the task, and reducing conflicts during transportation should be fully considered to reasonably set the number of AGVs and the AGV priorities, so that the entire scheduling system operates optimally. Therefore, the following scheduling objectives and principles are established:

[0141] Assume that the total freight volume to be transported this time is Q l , and the single-piece load of different types of goods is q r,i ={q r,1 ,q r ,2 ,...,q r,N}. The shipping volume of each trolley with different loading capacities is q li ={q l,1 ,q l,2 ,...,q l,N . The vehicles moving in the transportation direction are denoted as x s , and the empty return trolleys are denoted as x d , and their priorities are lx s ,lx d . The time window is (v i ,m i ), the time to complete the handling is t f , the number of loaded trolleys The total number of operating trolleys is m ω . The situation of conflicts is denoted as ▽ c

[0142] The objective is: t f ≤m i ;

[0143] The principles are as follows:

[0144] minc1 = mod(q l,i ,q r,i ); (1)

[0145] minc2 = minc s ; (2)

[0146]

[0147] lx s = lx d = l; if ▽ c (4)

[0148]

[0149] In the formula, it is expressed in turn as:

[0150] Principle 1: When retrieving idle trolleys from the ship docking point, according to the required carrying capacity q1 of the goods, give priority to arranging vehicles with a transportation capacity of q1·n (n = 1, 2, 3...), that is, vehicles whose AGV transportation capacity is a multiple of it for transportation. First, retrieve idle trolleys, and then retrieve trolleys in the charging state;

[0151] Principle 2: Since the scheduling of the ship has time window (v i , m i ) restrictions, it is necessary to consider the penalty cost c s generated by not completing the transportation within the specified time to be minimized;

[0152] Principle 3: When planning the path in the grid network between the starting point and the ending point, there will be a trolley k1 for load transportation from the starting point to the ending point, and a trolley k2 waiting to be rehandled for return from the ending point to the starting point. The vehicle transporting in the transportation direction is denoted as x s , and the empty return trolley is denoted as x d . When the number of trolleys with load tasks reaches half of the total number of operating trolleys, the priority of k1 is always greater than that of k2, and vice versa;

[0153] Principle 4: When there is a conflict in trolley transportation, according to the established priority function l, perform reasonable scheduling of vehicles:

[0154] Principle 5: When there is a task arrangement, by arranging the number of trolleys Q A , make the operation efficiency of the system better.

[0155] Generate the initial population of the sparrow optimization algorithm in step S4

[0156] Its corresponding fitness

[0157] In step S5, improve the identity transformation rule and position transformation rule of the sparrow optimization algorithm, including the following steps:

[0158] S5.1 Update the position of the discoverer:

[0159]

[0160] Where: t represents the current iteration number, j = 1, 2, 3...d. R2 ∈ [0, 1] and ST ∈ [0.5, 1] represent the early warning value and the safety value respectively. Q is a random number conforming to the normal distribution, L represents a 1×d matrix, where each element in the matrix is all 1. R2 < ST means less than the early warning range, and R2 ≥ ST means reaching the early warning range.

[0161] S5.2 Update the position of the joiner:

[0162]

[0163] Among them, X p represents the optimal position occupied by the current discoverer, and X worst represents the current global worst position. A is a 1×d matrix with all elements being 1 or -1, and A + = A T (AA T ) -1 .

[0164] S5.3 Update the position of the scout:

[0165]

[0166] Among them, X best is the current global optimal position. β is the step size control parameter and is a random number following a normal distribution with a mean of 0 and a variance of 1; f g and f w represent the current global best and worst fitnesses respectively. K is a random number within [-1, 1], and ε is a constant to avoid a zero denominator.

[0167] S5.4 Add the rule for the identity transformation of the joiner-discoverer:

[0168] Set the time t. When there is no better position under the set time t, a certain number m of joiners are directly changed to discoverers. Both t and m are dynamically changing;

[0169] Assume that the number of algorithm iterations is N and the current iteration number is n;

[0170] When it reaches the first , at this time, t follows At this time, m follows log n / N+1 n;

[0171] When it reaches the second , at this time, t follows At this time, m follows log N-n / N+1 n;

[0172] The meaning of this formula is that at the first , t becomes smaller and m becomes larger, but the growth rate of m decreases later.

[0173] S5.5 Add the rule for the position transformation of the joiner-discoverer:

[0174] When the discoverer finds food, there may be multiple joiners jumping towards its position simultaneously. This is not conducive to expanding the search efficiency in the initial stage. The flying speed of each joiner can be set so that when one joiner reaches the position of the discoverer, other sparrows are not allowed to fly over. The steps are as follows:

[0175] Step 1: Set the flying speeds of the joiners as v1, v2... v n , and the distances from the discoverer who found the food are d1, d2... d n ;

[0176] Step 2: When the discoverer finds food, compare the speeds v of each joiner k ;

[0177] Step 3: If the speed v of one of the joiners k makes d k / v k less than other values, then all other joiners maintain the original operation, and the joiner with this speed v k jumps to the position of the discoverer.

[0178] S5.6 proposes the relevant rules of the sparrow optimization algorithm regarding the visibility level:

[0179] When the number of times τ that a joiner changes its identity to a discoverer reaches a certain level, its eyesight will decline. The discoverer will not stop seeing. The advantage is that the visualization level is high in the early stage, expanding the search range, and the search range is reduced in the later stage when the eyesight declines, thus improving the convergence speed of the algorithm.

[0180] The specific implementation is as follows:

[0181] Step 1: Set the eyesight ε of the joining sparrows and the visibility level γ of the discoverer. The two are in a positive correlation, that is, ε = kγ (k > 0), the number of algorithm iterations is N, and the current iteration number is n;

[0182] Step 2: Judge whether the number of times σ that a joiner changes its identity to a discoverer reaches a certain number τ;

[0183] Step 3: If σ < τ, maintain the original search rule; if σ ≥ τ, ε changes accordingly at this time, the eyesight of the sparrow decreases, and the corresponding visibility level of the discoverer also decreases.

[0184] S5.7 judges whether the current iteration number n reaches the set maximum iteration number iter max :

[0185] if n > iter max , then output the value f of the best fitness at this time gAs the optimal objective value, the corresponding one is the optimal objective solution, so as to find out the best AGV scheduling scheme and the corresponding minimum value Z min ; Otherwise, continue to step S5.

[0186] The AGV scheduling method for automated terminals provided by the present invention makes up for the deficiencies in the current AGV scheduling scheme, extracts the delivery time of terminal goods, the node information of the shelves, and the vehicle operation information, etc., comprehensively considers the factor of the lowest empty-load rate, establishes a problem model regarding cost, time, and empty-load rate, considers factors such as penalty cost, and establishes a reasonable scheduling principle for multiple AGVs, which has important practical significance in practical problems; Secondly, this paper improves the sparrow optimization algorithm and applies it to the problem. The improvements include the position transformation rule and identity transformation rule of the joiner and discoverer, and the proposal to consider the vision and visualization degree of the constructed sparrows, so that the algorithm reduces the situation of converging to the local optimal solution, can not only quickly find the optimal solution, but also reasonably give the AGV path scheme.

[0187] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalent technologies, the present invention is also intended to include these modifications and variations.

Claims

1. A scheduling method for an automated terminal AGV, characterized in that: It includes the following steps: S1: Extract relevant information to obtain the path information for the transportation scheduling of the AGV in the automated terminal; S2: Establish the objective function and constraint conditions of the AGV scheduling method in the automated terminal according to the extracted node information; S3: Establish the overall AGV scheduling objectives and principles; S4: Use the sparrow optimization algorithm to solve the AGV scheduling problem, generate the initial population of the sparrow optimization algorithm, and calculate the population fitness; S5: Improve the identity transformation rule and position transformation rule of the sparrow optimization algorithm, and perform iterative operations until the current iteration number reaches the set maximum iteration number; S6: Output the best AGV scheduling plan and the optimal value of the corresponding objective function; The relevant information in step S1 includes: A{a1,a2,a3...a n} represents the set of AGVs, and H{h1,h2...h n} represents the set of shelf nodes. The congestion waiting time of an AGV K from shelf i to shelf j is denotes that an AGV K transports from shelf i to shelf j. represents the transportation time from shelf i to shelf j. represents the manual operation time at shelf i. (v i ,m i ) represents the time window of the goods docking shore. t g represents the time required for the trolley to charge after completing the transportation. t s represents the time generated by rescheduling the goods due to an accident. denotes an AGV K generates transportation costs during the transportation from i to j. represents the cost generated by manual operation at transportation point i. c s represents the penalty cost generated for not completing the transportation within the specified time; The objective function in step S2 includes: Objective function 1: This objective function is based on the requirement that when using AGV for arrival storage, the time required to install all goods is the shortest; Objective function 2: This objective function is based on minimizing the cost of completing the task, where: represents the weighting coefficient, where m q is the quantity of goods, and w q is the value of the goods attribute; Objective function 3: This objective function is based on arranging vehicles to perform multiple tasks simultaneously with the lowest vehicle empty load rate while meeting the time window constraints and its own capacity constraints, where: C is the capacity of an AGV vehicle, c u , u = 1, 2, 3... n, is the space capacity occupied by each task; The constraint conditions in step S2 include: c u ≤ C(5) Among them: Equation (1) indicates that the final time to complete all handling tasks does not exceed the maximum ship docking time; Equation (2) indicates whether the AGV is transporting on paths i and j; Equation (3) indicates that one task needs to be completed in one transportation; Equation (4) indicates that one container loading and unloading is completed by one AGV; Equation (5) represents the capacity constraint of the vehicle; In the step S3, the AGV scheduling target is t f ≤m i ; t f is the time to complete the handling The AGV scheduling principles are as follows: minc1 = mod(q l,i , q r,i ); (1) minc2 = minc s ; (2) where the total freight volume to be transported is Q l , the single-piece load of different types of goods is q r,i = {q r,1 , q r,2 ,..., q r ,N}, the shipment volume of each trolley with different loading capacities is q li = {q l,1 , q l,2 ,..., q l,N}, the vehicles moving in the transportation direction are denoted as x s , the empty trolleys returning are denoted as x d , then their priorities are lx s , lx d The time window is (v i , m i ), the number of loaded trolleys The total number of operating trolleys is m w , the situation of conflicts is denoted as t p represents the time required to complete a single freight transportation; In the formula, they are represented in turn as: (1) When retrieving idle vehicles from the ship docking point, according to the required carrying capacity q1 of the goods, preferentially arrange vehicles with a transportation capacity of q1·n for transportation, that is, the AGV transportation capacity is a multiple of it. First, retrieve idle vehicles, and then retrieve vehicles in the charging state; (2) Restricted by the time window (v i , m i ) of ship scheduling, considering the penalty cost c s generated for not completing the transportation within the specified time to be minimized; (3) When path planning is carried out in the grid network between the starting point and the ending point, there will be a trolley k1 for carrying loads from the starting point to the ending point, and a trolley k2 waiting to be reloaded and returning from the ending point to the starting point. The vehicle transporting in the transportation direction is denoted as x s , and the empty return trolley is denoted as x d , when the number of trolleys with load tasks reaches half of the total number of operating trolleys, the priority of k1 is always greater than that of k2, and vice versa; (4) When there is a conflict in vehicle transportation, perform reasonable vehicle scheduling according to the established priority function l: When there is a task arrangement, by arranging the number of trolleys Q A , to make the operation efficiency of the system better.

2. The scheduling method for an automated terminal AGV according to claim 1, characterized in that: The steps for the priority setting are as follows: Select two metrics. The first metric is the remaining order quantity q of the trolley j , and the second metric is the distance s to the shelf number j . The orders of the two metrics are l1 and l2 respectively. When congestion occurs, follow the following rules: Referring to the number m of trolleys in operation, the final priority rule l for trolley scheduling is Where M is the number of all operable vehicles, The meaning of this function is: The number of operating vehicles is less than At this time, according to l1, that is, the remaining order volume is prioritized; The number of running vehicles is more than According to L2, that is, the ones farther from the shelf have priority; The running quantity is between and According to the above rules, it mainly focuses on giving priority to scheduling based on the transportation volume.

3. The scheduling method of an automated terminal AGV according to claim 1, wherein: Generating the initial population of the sparrow optimization algorithm in step S4 Its corresponding fitness 4. The scheduling method of an automated terminal AGV according to claim 1, wherein: In step S5, improving the identity transformation rule and position transformation rule of the sparrow optimization algorithm includes the following steps: S5.1 Update the position of the discoverer: Among them: t represents the current iteration number, j = 1, 2, 3...d, R2∈[0, 1] and ST∈[0.5, 1] represent the warning value and safety value respectively, Q is a random number conforming to the normal distribution, L represents a 1×d matrix, where each element in the matrix is all 1, R2 < ST represents less than the warning range, and R2 ≥ ST represents reaching the warning range; S5.2 Update the position of the joiner: Among them, X p represents the optimal position occupied by the current discoverer, and X worst represents the current global worst position. A is a 1xd matrix with all elements being 1 or -1, and A + = A T (AA T ) -1 ; S5.3 Update the position of the scout: Where: X best is the current global optimal position, β is the step size control parameter, and is a random number that follows a normal distribution with a mean of 0 and a variance of 1; f g and f w represent the current global best and worst fitness respectively, K is a random number within [-1, 1], and ε is a constant to avoid a denominator of 0; S5.4 Add the identity transformation rule for the joiner and discoverer: Set the time t. When there is no better position at the set time t, directly change a certain number m of joiners to discoverers. Both t and m are dynamically changing; Assume that the number of algorithm iterations is N, and the current iteration number is n; When proceeding to the previous time, at this time t follows At this time m follows log n / N+1 n; When it comes to the later moment, at this time t follows At this time, m follows log N-n / N+1 n; Among them, before When it is, t becomes smaller and m becomes larger, but the growth rate of m decreases later; S5.5 Add the position transformation rule for the joiner and discoverer: When the discoverer discovers food, there will be a situation where multiple joiners jump to its position at the same time, which is not conducive to expanding the search efficiency in the initial stage. Set the flight speed of each joiner, so that when one reaches the joiner's position, other sparrows are not allowed to fly over; S5.6 Propose the relevant rules regarding the visibility degree of the sparrow optimization algorithm: When the number of times the joiner transforms into a discoverer reaches a certain number τ, the vision will decline. The discoverer will not become blind. The advantage is that the visualization level is high in the early stage, expanding the search range. In the later stage, the vision decline reduces the search range, thus increasing the convergence speed of the algorithm; S5.7 Determine whether the current iteration number n reaches the set maximum iteration number iter max If n > iter max then output the value f of the best fitness at this time g as the optimal objective value, and the corresponding one is the optimal objective solution, so as to find the best AGV scheduling scheme and the corresponding minimum value Z min ; Otherwise, continue with step S5.

5. The scheduling method of an automated terminal AGV according to claim 4, wherein: The specific steps of the said step S5.5 are as follows: Step 1: Set the flying speeds of the joiners as v1, v2... v n , and the distances from the discoverer who found the food as d1, d2... d n ; Step 2: When the discoverer finds food, compare the speed v of each joiner k ; Step 3: If the speed v of one of the joiners k , such that d k / v k is less than other values, then all other joiners maintain the original operation, and the joiner with this speed v k jumps to the discoverer's position.

6. The scheduling method of an automated terminal AGV according to claim 4, wherein: The specific steps of the said step S5.6 are as follows: Step 1: Set the vision ε of the joining sparrows and the visibility γ of the discoverers. The two are in a positive correlation, i.e., ε = kγ, where: The number of algorithm iterations is N, and the current iteration number is n; Step 2: Determine whether the number of times σ that the joiner transforms into a discoverer reaches a certain number τ; Step 3: If σ < τ, maintain the original search rule; if σ ≥ τ, ε changes accordingly at this time, the sparrow's vision decreases, and the corresponding visualization level of the discoverer also decreases.

Images