Intelligent scheduling method and system for vertical warehouse based on data analysis to obtain congestion coefficient

Abstract

This invention provides an intelligent scheduling method for automated storage and retrieval systems based on data analysis to obtain congestion coefficients, relating to the fields of data analysis and algorithm technology. The steps are as follows: S1: Establish a basic storage and retrieval model; S2: Obtain the congestion coefficient threshold and the historical storage and retrieval ratio and historical congestion coefficient corresponding to the current time, and adjust the status of the automated storage and retrieval system entrances and exits; S3: Obtain all storage and retrieval tasks, and determine whether to activate the simulated annealing algorithm to obtain the task permutation combination with the shortest required time; S4: If the simulated annealing algorithm is activated to obtain the task permutation combination with the shortest required time, output the task permutation combination and execute all storage and retrieval tasks in sequence; otherwise, directly execute all storage and retrieval tasks in the order they were obtained. This invention introduces a congestion coefficient into the automated storage and retrieval system scheduling system, which can effectively solve the congestion problem at the automated storage and retrieval system entrances and exits. Combined with the simulated annealing algorithm to obtain the task permutation combination with the shortest time, it reduces the waiting time for users during storage and retrieval.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data analysis and algorithm, in particular to a vertical warehouse intelligent scheduling method and system based on data analysis to obtain congestion coefficient. BACKGROUND

[0002] Due to the space-intensive characteristics of the vertical warehouse, it plays an important role in relieving the urban parking problem. Generally, due to the constraints of land and site conditions, for large-scale vertical warehouses, only multiple completely independent vertical warehouses can be established, and each warehouse has only a relatively small number of entrances and exits. The entrances and exits of the vertical warehouse will appear congestion phenomenon, causing the user's waiting time to be too long, and reducing the operation efficiency of the vertical warehouse.

[0003] In view of the above problems, the application number is: 202011195662.0, a data processing method based on artificial intelligence and intelligent parking lot and cloud server is disclosed, which obtains the parking lot operation log set and obtains the congestion area data set according to the parking lot operation log set, obtains the vehicle parking trajectory set based on the parking lot operation log set and obtains the vehicle parking reservation request set based on the congestion area data set, then obtains the parking space management data based on the vehicle parking trajectory set and the vehicle parking reservation request set, finally determines the parking space dynamic adjustment data of the parking lot operation log set according to the parking space management data and adjusts the number of parking spaces in the open state of the intelligent parking lot at different times based on the parking space adjustment data and the preset artificial intelligence model, which can to a certain extent avoid the congestion of the intelligent parking lot caused by a large number of vehicles entering the intelligent parking lot in a short time, but it still has the following shortcomings: adjusting the number of parking spaces in the open state by judging that there may be congestion, so that some parking spaces in the non-use state cannot work, to a certain extent, reducing the use rate of the vertical warehouse parking space, and it still cannot realize the function of making the user waiting time as short as possible when different users need to store and take the car at the same time. SUMMARY

[0004] The technical problem to be solved by the present application is how to solve the congestion problem existing in the entrances and exits of the vertical warehouse, and reduce the waiting time of the user when storing and taking.

[0005] To solve the above technical problems, the present application provides the following technical scheme: a vertical warehouse intelligent scheduling method based on data analysis to obtain congestion coefficient, comprising the following steps:

[0006] S1: establishing a basic access model;

[0007] S2: obtaining the congestion coefficient threshold value and the historical access ratio and historical congestion coefficient corresponding to the current time, and adjusting the state of the vertical warehouse entrance and exit;

[0008] S3: Obtain all access tasks, determine whether to start the simulated annealing algorithm to obtain the task permutation and combination with the shortest required time. The specific process of the simulated annealing algorithm is as follows:

[0009] S301: Obtain the set of access tasks to be assigned, A = {T} c1 ...T cm T q1 ...T qn}, where T c1 ...T cm To store the task, T q1 ...T qn To retrieve tasks; based on the different order of the stored tasks in the task set, a total of i possible task permutations and combinations are obtained: P1(A)...P i (A), SumT1...SumT i Given the time required for the corresponding task permutation and combination, the initial temperature is w0, w0 = min(2000, i), and the acceptable internal energy temperature is set to eps = 1.

[0010] S302: Generate an initial solution SumT x SumT x The time required to execute tasks directly in the order of retrieving all access tasks;

[0011] S303: Randomly swap the order of two tasks to generate a different solution (SumT). x ';

[0012] S304: Determine whether to accept the new solution SumT x ';

[0013] S305: Press w b+1 =w b *ΔT updates the temperature, where ΔT is the cooling coefficient and b is a natural number;

[0014] S306: Repeat steps S303-S305, when the temperature w b+1 When eps are reached, the simulated annealing algorithm ends, and the current task order is the order of the task combinations with the shortest time.

[0015] S307: Output the task permutation and combination that requires the shortest time;

[0016] S4: If the simulated annealing algorithm is started to obtain the task permutation and combination with the shortest time, then the task permutation and combination is output and all access tasks are executed in order. Otherwise, all access tasks are executed directly in the order in which they are obtained.

[0017] Preferably, in step S3, the specific process of determining whether to start the simulated annealing algorithm is as follows: if the historical congestion coefficient reaches the congestion coefficient threshold, then the simulated annealing algorithm is started; otherwise, it is not started.

[0018] Preferably, in step S3, if the simulated annealing algorithm is started and the calculation time exceeds 5 seconds, the simulated annealing algorithm ends immediately and outputs the task permutation and combination with the shortest current time.

[0019] Preferably, step S304 determines whether to accept the new solution SumT. x The specific process is as follows: If SumT x ' <SumT x Accept the new solution and update the current solution to SumT. x '; If SumT x '>=SumT x Then, based on probability Accept the new solution, where k is the Boltzmann constant and w is the current temperature.

[0020] Preferably, in step S1, the basic access model includes an elevator, a traversing trolley, a transporter, and a turntable; the movement time between adjacent floors of the elevator is t. d The movement time of adjacent columns of the transverse trolley is t. h The time for the transporter to pick up and place goods is t. b The rotation time of the turntable is t. z .

[0021] Preferably, the mathematical model for the time required for each access task in the basic access model is as follows:

[0022] When the elevator is at the starting position, O(i0, j0) represents the entrance / exit currently operated by the user, and Q1(i1, j1) represents the coordinates of the target storage location.

[0023] The time required for a single inventory task is t. c =2t b +|i1-i0|*t d +|j1-j0|*t h ;

[0024] The time required for a single pickup is t. q =2t b +max(|i1-i0|*t d ,t z )+2|j1-j0|*t h ;

[0025] When the elevator is not at the starting position but at Q2(i2, j2), O(i0, j0) is the entrance / exit currently operated by the user, and Q1(i1, j1) is the coordinate of the target storage location.

[0026] The time required for a single inventory check is t. c ′=|i2-i0|*t d +|j2-j0|*t h +2t b +|i1-i0|*t d +|j1-j0|*t h ;

[0027] The time required for a single pickup is t. q ′=|i2-i1|*t d +2|j2-j1|*t h +2t b +max(|i1-i0|*t d ,t z )+|j2-j0|*t h .

[0028] Preferably, in step S2, the specific process of adjusting the status of the automated warehouse entrances and exits is as follows: determine whether the historical congestion coefficient has reached the threshold; if the congestion coefficient has reached the threshold, set the entrances and exits with the corresponding historical storage and retrieval ratio as storage entrances or retrieval entrances respectively, based on the historical storage and retrieval ratio at the corresponding time; if the congestion coefficient has not reached the threshold, set all entrances and exits to a state that allows both storage and retrieval.

[0029] Preferably, the formula for calculating the congestion coefficient D is: Where B is the maximum number of goods passing through the entrance / exit per unit time, a is the actual number of goods passing through per unit time at the current moment, c is the capacity of the automated storage system, Δt is the difference between the current average storage / retrieval time and the historical average storage / retrieval time, and r and l are dynamic adjustment coefficients.

[0030] Preferably, the period for obtaining historical access ratios and historical congestion coefficients is from Monday to Sunday, with a minimum acquisition time of 15 minutes.

[0031] Corresponding to the above method, the present invention also provides an intelligent scheduling system for automated warehouses based on data analysis to obtain congestion coefficients, comprising the following modules:

[0032] Modeling module: Used to build the basic access model;

[0033] The module for adjusting the status of entrances and exits is used to obtain the congestion coefficient threshold and the historical data corresponding to the current moment, including the storage and retrieval ratio and the congestion coefficient, and to adjust the status of the entrances and exits of the automated warehouse.

[0034] The module for obtaining the optimal task permutation and combination is used to obtain all access tasks and determine whether to start the simulated annealing algorithm to obtain the task permutation and combination with the shortest required time.

[0035] The task execution module is used to output the task permutation and combination after the module for obtaining the optimal task permutation and combination has finished working, and to execute all access tasks in sequence. If the module for obtaining the optimal task permutation and combination is not started, all access tasks are executed directly in the order of obtaining access tasks.

[0036] The advantages of this invention are:

[0037] (1) Based on historical data, including historical access ratio and historical congestion coefficient, this invention predicts access demand and congestion at different time periods and dynamically adjusts the status of the vertical warehouse entrance and exit, which can effectively solve the problem of congestion at the vertical warehouse entrance and exit during peak periods.

[0038] (2) Based on the different historical congestion coefficients, determine whether the simulated annealing algorithm should be started. If congestion is detected, the simulated annealing algorithm is started. This can minimize the time required to complete all access tasks while resolving congestion, thus reducing user waiting time. If no congestion is detected, the algorithm is not started, thus avoiding waste of resources.

[0039] (3) The present invention can also flexibly adjust the solution range and convergence speed of the simulated annealing algorithm to avoid the simulated annealing algorithm from taking too long to calculate, thereby further ensuring that the user's waiting time is reduced and the user experience is improved. Attached Figure Description

[0040] Figure 1 This is a flowchart of the entirety of Embodiment 1 of the present invention;

[0041] Figure 2 This is a congestion prediction map from Embodiment 1 of the present invention;

[0042] Figure 3 This is a flowchart illustrating the adjustment of the entrance / exit status in Embodiment 1 of the present invention. Detailed Implementation

[0043] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below in conjunction with the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0044] Example 1

[0045] like Figure 1The diagram shown is a flowchart of the overall implementation of Embodiment 1 of the present invention. The present invention provides a method for intelligent scheduling of automated warehouses based on data analysis to obtain congestion coefficients, including the following steps:

[0046] S1: Establish the basic access model; the specific modeling process is as follows:

[0047] S101: Establish a correspondence between each storage location using a coordinate matrix. For example, (2,3) represents storage location 3 on floor 2. Let the travel time between adjacent floors of the elevator be t. d The time between adjacent movements of the transverse trolley is t. h The time for the transporter to pick up and place goods is t. b The rotation time of the turntable is t. z Because the lifting and turning of the elevator and the rotation of the turntable are currently synchronized, the maximum value of the two times is taken.

[0048] S102: Determine the initial position of the elevator;

[0049] S103: If the elevator is initially located at the starting point, i.e., initially located at the entrance / exit, let O(i0, j0) be the entrance / exit currently operated by the user, and Q1(i1, j1) be the coordinates of the target storage location.

[0050] t c Let t be the time required for a single inventory transaction. The specific inventory process is as follows: the transporter moves the goods onto the elevator, the elevator rises to the floor where Q1(i1, j1) is located, after reaching the target floor, the traverse trolley moves laterally to the target storage location, and the transporter moves the goods to the target storage location. The mathematical model for this inventory process is: t c =2t b +|i1-i0|*t d +|j1-j0|*t h ,

[0051] t q The time required for a single pickup is t. The specific pickup process is as follows: Before the traverse trolley moves to the target storage location, the transporter moves the goods onto the elevator. The elevator begins to descend, and the turntable rotates simultaneously during the descent. After the elevator descends to the floor where the entrance / exit is located, the traverse trolley moves to the entrance / exit position, and the transporter moves the goods to the entrance / exit storage location. The mathematical model for this pickup process is: t q =2t b +max(|i1-i0|*t d ,t z )+2|j1-j0|*t h ;

[0052] If the elevator is initially located at Q2(i2, j2) instead of the starting position, let O(i0, j0) be the entrance / exit currently operated by the user, and Q1(i1, j1) be the coordinates of the target storage location.

[0053] t c 'Takes the time required for a single inventory transaction. The specific inventory process is as follows: The elevator descends to the entrance / exit level, the traverse trolley moves laterally to the entrance / exit position, the transporter moves the goods onto the elevator, the elevator rises to the floor where Q1(i1, j1) is located, after reaching the target floor, the traverse trolley moves laterally to the target storage location, and the transporter moves the goods to the target storage location. The mathematical model for this inventory process is: t c ′=|i2-i0|*t d +|j2-j0|*t h +2t b +|i1-i0|*t d +|j1-j0|*t h ,

[0054] t q 'T0 represents the time required for a single pickup. The specific pickup process is as follows: the elevator rises to the target cargo level, the traverse trolley moves laterally to the target storage location, the transporter moves the cargo onto the elevator, the elevator begins to descend, and the turntable rotates simultaneously during the descent. After the elevator reaches the floor where the entrance / exit is located, the traverse trolley moves laterally to the entrance / exit position, and the transporter moves the cargo to the entrance / exit storage location. The mathematical model for this pickup process is: t0 q ′=|i2-i1|*t d +2|j2-j1|*t h +2t b +max(|i1-i0|*t d ,t z )+|j2-j0|*t h .

[0055] S2: Using Monday to Sunday as a statistical period, with 15 minutes as the minimum statistical time period, there are 96 time periods per day. The access ratio and congestion coefficient of the corresponding time periods are statistically analyzed as historical data, and the congestion coefficient threshold D is calculated. Where B is the maximum throughput capacity of the entrance / exit, i.e. the number of goods that can pass through the entrance / exit per unit time, a is the actual number of goods that pass through per unit time at the current moment, c is the capacity of the automated storage and retrieval system, i.e. the total number of goods that the automated storage and retrieval system can hold, Δt is the difference between the current average storage and retrieval time and the historical average storage and retrieval time, and r and l are dynamic adjustment coefficients.

[0056] like Figure 2As shown, by collecting historical and real-time data on the operation of the automated storage system, cleaning and preparing the data, handling missing values ​​and outliers, and analyzing and organizing the data, a congestion prediction map is obtained. Based on this prediction map, the congestion coefficient and access status for each time period of the day can be predicted, and special needs during peak hours can be addressed in a timely manner.

[0057] like Figure 3 As shown, after obtaining the congestion coefficient threshold and the historical storage / retrieval ratio and historical congestion coefficient at the current moment, the status of the automated warehouse entrances / exits is adjusted. The specific process for adjusting the status of the automated warehouse entrances / exits is as follows: First, obtain the historical storage / retrieval ratio and historical congestion coefficient at the current moment, and then determine whether the historical congestion coefficient has reached the threshold; if the congestion coefficient has reached the threshold, set the entrances / exits with the corresponding storage / retrieval ratio as storage ports or retrieval ports respectively, based on the historical storage / retrieval ratio at the corresponding moment; if the congestion coefficient has not reached the threshold, set all entrances / exits to a state that allows both storage and retrieval.

[0058] S3: After the entrance and exit of the storage warehouse are adjusted to the corresponding states, all storage and retrieval tasks are retrieved. If the congestion coefficient reaches the threshold, the SA algorithm (specifically, the simulated annealing algorithm) is activated, and the basic storage and retrieval model is called to obtain the task permutation and combination with the shortest required time. Otherwise, the algorithm is not activated. The specific process of using the simulated annealing algorithm to obtain the task permutation and combination with the shortest required time is as follows:

[0059] S301: Obtain the set of access tasks to be assigned, A = {T} c1 ...T cm T q1 ...T qn}, where T c1 ...T cm For inventory tasks, t c1 ...t cm T represents the time required for each inventory task. q1 ...T qn For the pickup task, t q1 ...t qn Let P be the time required for each pickup task; based on the different order of the pickup and drop-off tasks in the task set, a total of i possible task permutations are obtained: P1(A)...P i (A), T1...T i To calculate the time required to complete all access tasks in the corresponding task permutations in sequence, the access time mathematical model in the basic access model is called to calculate the time required for each task. Given an initial temperature w0, w0 = min(2000, i), the acceptable internal energy temperature eps = 1 is set.

[0060] S302: Generate an initial solution SumT x SumTx The time required to execute tasks directly in the order of retrieving all access tasks;

[0061] S303: Randomly swap the order of two tasks to generate a different solution (SumT). x At this point, different task orders yield different solutions (SumT). x The reason is:

[0062] The elevator stops at the target location after the inventory task is completed, and stops at the entrance / exit level after the retrieval task is completed. In other words, the movement trajectory of subsequent tasks is determined.

[0063] Assume there is only one entrance / exit currently in operation, and there are three tasks {T}. c1 T c2 T q1 The process is queued and executed, and the required storage and retrieval time is calculated using the mathematical model from the basic storage and retrieval model. c1 =110s,t c2 =150s,t q1 =90s;

[0064] Because tasks are executed sequentially, each task must wait for the preceding task to finish before it can execute. Therefore, the waiting time for the user varies depending on the task order, and the sum of the times from task creation to completion also varies. Specifically:

[0065] When the order of task execution is T c1 T c2 T q1 At that time, the sum of the execution times of each task is: SumT x =t c1 +(t c1 +t c2 )+(t c1 +t c2 +t q1 ) = 3t c1 +2t c2 +t q1 =720s;

[0066] The order of task execution is changed to T. q1 T c1 T c2 At that time, the sum of the execution times of each task is: SumT x '=t q1 +(t q1 +t c1 )+(t q1 +t c1 +t c2 ) = 3t q1 +2t c1 +tc2 =640s;

[0067] That is, by optimizing the task execution order, the total time spent by each user is calculated, so that the overall time spent by the user group is less;

[0068] Two tasks may also be executed simultaneously at different entrances / exits. Let's assume there are still three tasks {T}. c1 T c2 T q1} Queued execution, but T c2 relative to T c1 T q1 For tasks executed concurrently at independent entry and exit points, i.e., with T c2 With T c1 Execute simultaneously, at which point T is executed. c2 The waiting time is 0;

[0069] When the order of task execution is T c1 T c2 T q1 At that time, T c1 After execution, T q1 Execute immediately, no need to wait for T c2 Execution complete, the sum of the execution times of all tasks is: SumT x =t c1 +t c2 +(t c1 +t q1 ) = 2t c1 +t c2 +t q1 =460s

[0070] The order of task execution is changed to T. q1 T c1 T c2 At that time, T q1 and T c2 Execute simultaneously, T q1 After execution, T c2 Execute immediately, no need to wait for T c1 Execution complete, the sum of the execution times of all tasks is: SumT x '=t q1 +t c1 +(t q1 +t c2 ) = 2t q1 +t c1 +t c2 =440s;

[0071] Therefore, changing the order of tasks will result in different sums of the execution times of all tasks.

[0072] S304: Determine whether to accept the new solution SumT x ', if SumT x ' <SumT x Accept the new solution and update the current solution to SumT. x '; If SumT x '>=SumT x Then, based on probability Accept the new solution, where k is the Boltzmann constant and w is the current temperature;

[0073] S305: Press w b+1 =w b *ΔT updates the temperature, where ΔT is the cooling coefficient and b is a natural number;

[0074] S306: Repeat steps S303-S305, when the temperature w b+1 When eps are reached, the simulated annealing algorithm ends, and the current task order is the order of the task combinations with the shortest time.

[0075] S307: Output the task permutation and combination that requires the shortest time;

[0076] In this embodiment, when starting the simulated annealing algorithm, a dynamic convergence method is also incorporated. That is, when the computation time of the simulated annealing algorithm exceeds 5 seconds, the algorithm immediately terminates and outputs the task permutation and combination with the shortest current time.

[0077] S4: If the simulated annealing algorithm is started to obtain the task permutation and combination with the shortest required time, then the task permutation and combination is output and all access tasks are executed in order. If the simulated annealing algorithm is not started, then all access tasks are executed directly in the order in which they are obtained.

[0078] This embodiment statistically analyzes storage and retrieval data from two dimensions: storage and retrieval ratio and congestion coefficient. It dynamically adjusts the status of the automated storage and retrieval system's entrances and exits, enabling more scientific utilization of storage space resources, improving the operational efficiency and resource utilization of the automated storage and retrieval system, effectively solving the problem of congestion at the entrances and exits of the automated storage and retrieval system during peak periods, reducing user waiting time, and providing users with better storage and retrieval services.

[0079] It also predicts the situation at different times based on different congestion coefficients, and then determines whether to start the simulated annealing algorithm. Its core idea is to accept a worse solution with a certain probability to avoid getting trapped in local optima too early, and to have a higher probability of finding the global optimum. While avoiding congestion, it can also complete all access tasks in the shortest possible time. In addition, it combines dynamic convergence method to flexibly adjust the solution range and convergence speed of the algorithm, further ensuring that user waiting time can be reduced. If it is determined that there will be no congestion, the simulated annealing algorithm will not be started to avoid unnecessary waste of resources.

[0080] Example 2

[0081] Corresponding to Embodiment 1 of the present invention, Embodiment 2 of the present invention also provides an intelligent scheduling system for automated warehouses based on data analysis to obtain congestion coefficients, comprising the following modules:

[0082] Modeling module: Used to build the basic access model, specifically including the following units:

[0083] Establish a coordinate system unit: This is used to establish a one-to-one correspondence between each storage location using a coordinate matrix. For example, (2,3) represents storage location 3 on floor 2. Let the travel time between adjacent floors of the elevator be t. d The time between adjacent movements of the transverse trolley is t. h The time for the transporter to pick up and place goods is t. b The rotation time of the turntable is t. z Because the lifting and turning of the elevator and the rotation of the turntable are currently synchronized, the maximum value of the two times is taken.

[0084] Elevator position determination unit: used to determine the initial position of the elevator;

[0085] Establish a mathematical model unit: used to calculate the time required for storage and retrieval. The specific calculation method is as follows:

[0086] If the elevator is initially located at the starting position, i.e., at the entrance / exit, let O(i0, j0) be the entrance / exit currently operated by the user, and Q1(i1, j1) be the coordinates of the target storage location.

[0087] t c Let t be the time required for a single inventory transaction. The specific inventory process is as follows: the transporter moves the goods onto the elevator, the elevator rises to the floor where Q1(i1, j1) is located, after reaching the target floor, the traverse trolley moves laterally to the target storage location, and the transporter moves the goods to the target storage location. The mathematical model for this inventory process is: t c =2t b +|i1-i0|*t d +|j1-j0|*t h ,

[0088] t q The time required for a single pickup is t. The specific pickup process is as follows: the traverse trolley moves laterally to the target storage location, the transporter loads the goods onto the elevator, the elevator begins to descend, and the turntable rotates simultaneously during the descent. After the elevator descends to the floor where the entrance / exit is located, the traverse trolley moves laterally to the entrance / exit position, and the transporter moves the goods to the entrance / exit storage location. The mathematical model for this pickup process is: t q =2t b +max(|i1-i0|*t d ,t z)+2|j1-j0|*t h ;

[0089] If the elevator is initially located at Q2(i2, j2) instead of the starting position, let O(i0, j0) be the entrance / exit currently operated by the user, and Q1(i1, j1) be the coordinates of the target storage location.

[0090] t c 'Takes the time required for a single inventory transaction. The specific inventory process is as follows: The elevator descends to the entrance / exit level, the traverse trolley moves laterally to the entrance / exit position, the transporter moves the goods onto the elevator, the elevator rises to the floor where Q1(i1, j1) is located, after reaching the target floor, the traverse trolley moves laterally to the target storage location, and the transporter moves the goods to the target storage location. The mathematical model for this inventory process is: t c ′=|i2-i0|*t d +|j2-j0|*t h +2t b +|i1-i0|*t d +|j1-j0|*t h ,

[0091] t q 'T0 represents the time required for a single pickup. The specific pickup process is as follows: the elevator rises to the target cargo level, the traverse trolley moves laterally to the target storage location, the transporter moves the cargo onto the elevator, the elevator begins to descend, and the turntable rotates simultaneously during the descent. After the elevator reaches the floor where the entrance / exit is located, the traverse trolley moves laterally to the entrance / exit position, and the transporter moves the cargo to the entrance / exit storage location. The mathematical model for this pickup process is: t0 q ′=|i2-i1|*t d +2|j2-j1|*t h +2t b +max(|i1-i0|*t d ,t z )+|j2-j0|*t h .

[0092] The module for adjusting the status of entrances and exits is used to obtain the congestion coefficient threshold and the historical data corresponding to the current moment, including the storage and retrieval ratio and the congestion coefficient, and to adjust the status of the entrances and exits of the automated warehouse. The specific adjustment process is as follows: First, obtain the historical storage and retrieval ratio and the historical congestion coefficient corresponding to the current moment. Then, determine whether the historical congestion coefficient has reached the threshold. If the congestion coefficient has reached the threshold, set the entrances and exits with the corresponding storage and retrieval ratio as storage ports or retrieval ports, respectively, according to the historical storage and retrieval ratio at the corresponding moment. If the congestion coefficient has not reached the threshold, set all entrances and exits to a state that allows both storage and retrieval.

[0093] The module for obtaining the optimal task permutation and combination is used to activate the simulated annealing algorithm unit to obtain the task permutation and combination with the shortest required time when the congestion coefficient reaches the congestion coefficient threshold. The simulated annealing algorithm unit performs the following steps: First, it obtains the set of access tasks to be assigned, A = {T}. c1 ...T cm T q1 ...T qn}, where T c1 ...T cm For inventory tasks, T q1 ...T qn For the pickup task; based on the different order of the pickup and drop-off tasks in the task set, a total of i possible task permutations and combinations are obtained:

[0094] P1(A)...P i (A), T1...T i Given the time required for the corresponding task permutations and combinations, an initial temperature w0, w0 = min(2000, i), and an acceptable internal energy temperature eps = 1, generate an initial solution SumT. x SumT x The time required to execute tasks directly in the order of retrieving all access tasks; randomly swapping the order of two tasks generates a different solution, SumT. x Then determine whether to accept the new solution SumT. x ', if SumT x ' <SumT x Accept the new solution and update the current solution to SumT. x '; If SumT x '>=SumT x Then, based on probability Accept the new solution, where k is the Boltzmann constant and w is the current temperature; according to w b+1 =w b *ΔT updates the temperature, where ΔT is the cooling coefficient and b is a natural number; repeatedly and randomly change the order of the two tasks and determine whether to accept the new solution, when the temperature w b+1 When the time eps is reached, the simulated annealing algorithm ends. The current task order is the order of the task permutations with the shortest time. Finally, the task permutations with the shortest time are output.

[0095] The task execution module is used to output the task permutation and combination after the module for obtaining the optimal task permutation and combination has finished working, and to execute all access tasks in sequence. If the module for obtaining the optimal task permutation and combination is not started, all access tasks are executed directly in the order of obtaining access tasks.

[0096] The mathematical model building unit in the modeling module of this embodiment can calculate the required time for each access task in real time when the optimal task permutation and combination module is working, i.e., when the congestion coefficient reaches the congestion coefficient threshold and it is determined that congestion may occur. This is done in conjunction with the optimal task permutation and combination module to obtain the task permutation and combination with the shortest required time. Combined with the entrance / exit status adjustment module and the task execution module, this can effectively solve the problem of congestion during peak hours of the automated storage system, improve the operating efficiency and resource utilization of the automated storage system, reduce user waiting time, and enhance user experience. If the congestion coefficient does not reach the threshold, i.e., it is determined that no congestion will occur, the optimal task permutation and combination module will not work, thus avoiding waste of resources.

[0097] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for intelligent scheduling of automated warehouses based on data analysis to obtain congestion coefficients, characterized in that: Includes the following steps: S1: Establish a basic access model; the basic access model includes an elevator, a traversing trolley, a transporter, and a turntable; the movement time between adjacent floors of the elevator is... t d The movement time of adjacent columns of the transverse trolley is t h The time for the transporter to pick up and place goods is... t b The rotation time of the turntable is t z ; S2: Obtain the congestion coefficient threshold and the historical access ratio and historical congestion coefficient corresponding to the current moment, and adjust the status of the automated warehouse entrance and exit; the congestion coefficient D The calculation formula is: D= ,in B This represents the maximum number of goods that can pass through the entrance / exit per unit of time. a This represents the number of goods actually passing through per unit of time at the current moment. c For the capacity of the vertical storage, This is the difference between the current average access time and the historical average access time. r , l This is a dynamic adjustment coefficient; The specific process for adjusting the status of the automated warehouse entrances and exits is as follows: determine whether the historical congestion coefficient has reached the threshold; if the congestion coefficient has reached the threshold, set the entrances and exits with the corresponding historical storage and retrieval ratio as storage entrances or retrieval entrances respectively, based on the historical storage and retrieval ratio at the corresponding time; if the congestion coefficient has not reached the threshold, set all entrances and exits to a state that allows both storage and retrieval. S3: Obtain all access tasks, determine whether to start the simulated annealing algorithm to obtain the task permutation and combination with the shortest required time. The specific process of the simulated annealing algorithm is as follows: S301: Obtain the set of access tasks to be assigned. A= { T c1 ...T cm ，T q1 ...T qn },in T c1 ...T cm For inventory tasks, T q1 ...T qn For the pickup task; based on the different order of the pickup and drop-off tasks in the task set, the total is obtained. i Various task permutations and combinations: P 1 (A)...P i (A) , SumT 1... SumT i Given the time required for the corresponding task permutations and combinations, and an initial temperature. w 0, w 0= min (2000, i ), set the acceptable temperature of internal energy. eps =1; S302: Generate an initial solution SumT x , SumT x The time required to execute tasks directly in the order of retrieving all access tasks; S303: Randomly swap the order of two tasks to generate a different solution. SumT x ’ ; S304: Determine whether to accept the new solution SumT x ’ ; S305: Press Update temperature. The cooling coefficient, b It is a natural number; S306: Repeat steps S303-S305, when the temperature... achieve eps When the simulated annealing algorithm ends, the current task order is the order of the task combinations with the shortest required time. S307: Output the task permutation and combination that requires the shortest time; S4: If the simulated annealing algorithm is started to obtain the task permutation and combination with the shortest time, then the task permutation and combination is output and all access tasks are executed in order. Otherwise, all access tasks are executed directly in the order in which they are obtained.

2. The intelligent scheduling method for automated warehouses based on data analysis to obtain congestion coefficients as described in claim 1, characterized in that, In step S3, the specific process of determining whether to start the simulated annealing algorithm is as follows: if the historical congestion coefficient reaches the congestion coefficient threshold, then start the simulated annealing algorithm; otherwise, do not start it.

3. The intelligent scheduling method for automated warehouses based on data analysis to obtain congestion coefficients as described in claim 2, characterized in that, In step S3, if the simulated annealing algorithm is started and the calculation time exceeds 5 seconds, the simulated annealing algorithm ends immediately and outputs the task permutation and combination with the shortest current time.

4. The intelligent scheduling method for automated warehouses based on data analysis to obtain congestion coefficients according to claim 1, characterized in that, Step S304 determines whether to accept the new solution. SumT x ’ The specific process is as follows: If SumT x ’ < SumT x Accept the new solution and update the current solution. SumT x ’ ;like SumT x ’ >= SumT x Then, based on probability Accept the new solution, in which k Boltzmann's constant, w This is the current temperature.

5. The intelligent scheduling method for automated warehouses based on data analysis to obtain congestion coefficients according to claim 1, characterized in that, The mathematical model for the time required for each access task in the basic access model is as follows: When the elevator is at the starting position O( ， ) The entry and exit points currently being used by the user. Q 1 ( ， ) The coordinates of the target storage location. The time required for a single inventory check is ； The time required for a single pickup is ; When the elevator is not at the starting position but at... Q 2 ( ， ) hour, O( ， ) The entry and exit points currently being used by the user. Q 1 ( ， ) The coordinates of the target storage location. The time required for a single inventory check is ; The time required for a single pickup is .

6. The intelligent scheduling method for automated warehouses based on data analysis to obtain congestion coefficients according to claim 1, characterized in that, The historical access ratio and historical congestion coefficient are obtained from Monday to Sunday, with a minimum acquisition time of 15 minutes.

7. A smart warehouse scheduling system based on data analysis to obtain congestion coefficients, characterized in that, Includes the following modules: Modeling module: used to establish the basic access model; the basic access model includes an elevator, a traversing trolley, a transporter, and a turntable; the movement time between adjacent floors of the elevator is... t d The movement time of adjacent columns of the transverse trolley is t h The time for the transporter to pick up and place goods is... t b The rotation time of the turntable is t z ; The module for adjusting the status of entrances and exits is used to obtain the congestion coefficient threshold and the historical data corresponding to the current moment, including the storage and retrieval ratio and the congestion coefficient, and to adjust the status of the entrances and exits of the automated warehouse. The congestion coefficient D The calculation formula is: D= ,in B This represents the maximum number of goods that can pass through the entrance / exit per unit of time. a This represents the number of goods actually passing through per unit of time at the current moment. c For the capacity of the vertical storage, This is the difference between the current average access time and the historical average access time. r , l This is a dynamic adjustment coefficient; The specific process for adjusting the status of the automated warehouse entrances and exits is as follows: determine whether the historical congestion coefficient has reached the threshold; if the congestion coefficient has reached the threshold, set the entrances and exits with the corresponding historical storage and retrieval ratio as storage entrances or retrieval entrances respectively, based on the historical storage and retrieval ratio at the corresponding time; if the congestion coefficient has not reached the threshold, set all entrances and exits to a state that allows both storage and retrieval. The module for obtaining the optimal task permutation and combination is used to obtain all access tasks and determine whether to start the simulated annealing algorithm to obtain the task permutation and combination with the shortest required time. The task execution module is used to output the task permutation and combination after the module for obtaining the optimal task permutation and combination has finished working, and to execute all access tasks in sequence. If the module for obtaining the optimal task permutation and combination is not started, all access tasks are executed directly in the order of obtaining access tasks.

Images