CN116873431B - A multi-load AGV storage and transportation method based on slate intelligent warehouse - Google Patents

Abstract

The invention discloses a multi-heavy-load AGV storage and transportation method based on a rock plate intelligent warehouse, which comprises the following steps: constructing a cargo space allocation algorithm; constructing a heavy load AGV scheduling rule algorithm; generating a plurality of combined optimization solving algorithms through a composite algorithm framework according to a cargo space distribution algorithm and a heavy load AGV scheduling rule algorithm; constructing a simulation platform; determining a plurality of simulation calculation examples, and performing simulation solution on each simulation calculation example by adopting a plurality of combined optimization solution algorithms in a simulation platform to obtain a plurality of simulation results; analyzing the different performance evaluation indexes of each simulation result to determine a combined optimization solving algorithm which performs best under the different performance evaluation indexes of each simulation result; and correspondingly generating a storage and transportation strategy of the multi-load AGV. The invention solves the problems that the existing multi-heavy-load AGV intelligent warehouse is unfavorable for the management of the rock plate intelligent warehouse because goods come in and go out of the warehouse due to unreasonable coordination of goods space distribution and AGV dispatching.

Description

A multi-load AGV storage and transportation method based on slate intelligent warehouse

Technical field

The present invention relates to the technical field of multi-load AGV storage and transportation, in particular to a multi-load AGV storage and transportation method based on a slate intelligent warehouse.

Background technique

As my country rapidly strengthens the transformation of warehousing links from automation to intelligence, the transformation of the warehousing and logistics industries has put forward higher demands for intelligence. With the advancement of intelligent development, transportation equipment in warehousing operations has become intelligent. Automated Guided Vehicles (AGVs) have been introduced to replace forklifts, realizing unmanned operations.

The storage and transportation problem of multi-load AGVs in a slate smart warehouse can be described as: there are multiple AGVs and multiple transportation tasks in a slate smart warehouse. The transportation tasks are divided into inbound transportation tasks and outbound transportation tasks. The transportation tasks entering the warehouse come from the workshop production line and arrive randomly according to a uniform distribution. The outbound transportation tasks come from orders placed by customers, and their arrival times also have a certain degree of randomness. When the warehousing transportation task reaches the docking station, the rock slab storage strategy arranges appropriate cargo spaces for the arriving rock slabs and needs to meet the storage constraints of the goods and standard cargo spaces. After the cargo location of the warehousing transportation task is determined, the AGV scheduling strategy schedules the idle AGV, and the scheduled AGV goes to the docking station to store the warehousing task and complete the warehousing transportation task. When it reaches the outbound time, as the customer order is placed, the AGV scheduling strategy schedules the idle trolleys, and the scheduled AGV unloads the rock slabs from the storage space according to the customer's order, which completes the outbound transportation task.

There is a spatio-temporal imbalance in the warehousing tasks and outbound tasks of the existing multi-load AGV smart warehouse, that is, the warehousing tasks are carried out during the day, and the orders are shipped out at night, resulting in a long warehousing time and a short outgoing time. . Due to unreasonable cooperation between cargo space allocation and AGV scheduling, goods often have no time to enter and exit the warehouse, which is not conducive to the management of slate smart warehouses.

Contents of the invention

In view of the above defects, the present invention proposes a multi-load AGV storage and transportation method based on a slate intelligent warehouse. Its purpose is to solve the spatio-temporal imbalance in the warehousing tasks and out-of-warehouse tasks of the existing multi-load AGV intelligent warehouse. Moreover, due to the unreasonable cooperation between cargo space allocation and AGV scheduling, the goods are not able to enter and exit the warehouse in time, which is not conducive to the management of the slate smart warehouse.

To achieve this goal, the present invention adopts the following technical solutions:

A multi-load AGV storage and transportation method based on slate intelligent warehouse includes the following steps:

Step S1: Construct a cargo space allocation algorithm. The cargo allocation algorithm includes a cargo space allocation algorithm related to the frequency of goods leaving the warehouse, a cargo space allocation algorithm based on the value of the goods, a random cargo space allocation algorithm, and the cargo space allocation closest to the connection point. Algorithm and cargo space allocation algorithm in intermediate storage areas;

Step S2: Construct a heavy-load AGV scheduling rule algorithm. The heavy-load AGV scheduling rule algorithm includes a first-come-first-serve rule algorithm, a nearest rule algorithm, a furthest rule algorithm, a highest idle rule algorithm, a shortest running distance rule algorithm, and a random rule algorithm. ;

Step S3: According to the cargo space allocation algorithm and the heavy-load AGV dispatching rule algorithm, generate multiple combination optimization solution algorithms through a composite algorithm framework;

Step S4: Build a simulation platform;

Step S5: Determine multiple simulation examples, use multiple combination optimization solving algorithms in the simulation platform to simulate and solve each of the simulation examples, and obtain multiple simulation results;

Step S6: Analyze different performance evaluation indicators for each of the simulation results to determine the best combined optimization solution algorithm for each of the simulation results under different performance evaluation indicators;

Step S7: Based on the best-performing combined optimization solution algorithm under different performance evaluation indicators of each of the simulation results, correspondingly generate a storage and transportation strategy for the multi-load AGV.

Preferably, in step S1, before constructing the cargo space allocation algorithm, the following steps are also included:

Step S11: Determine the cargo encoding method;

Step S12: Use the cargo coding method to code the cargo locations in the smart warehouse;

Step S13: Divide the status of the coded goods.

Preferably, in step S1, in the cargo space allocation algorithm associated with the frequency of goods leaving the warehouse, the calculation formula for the frequency of goods leaving the warehouse is as follows:

Among them, GI _i represents the outbound frequency of goods of category i, i=1, 2,...I; GN _i represents the outbound quantity of goods of category i in the past period;

In the cargo location allocation algorithm based on the value of the goods, the value calculation formula of the goods is as follows:

Among them, GVI _i represents the value coefficient of goods of type i, i=1, 2,...I; G _i represents the specifications of goods of type i, expressed by the number of standard cargo spaces occupied by the goods; GN _i represents the number of standard goods of type i The quantity of goods shipped out of the warehouse in the past period.

Preferably, in step S5, before determining multiple simulation examples, the following steps are also included:

Determine the influencing factors of the simulation example. The influencing factors of the simulation example include AGV speed in no-load state, AGV speed in loaded state, connection station capacity, order arrival time interval and task type ratio.

Preferably, in step S6, the determination of the best performing combinatorial optimization solution algorithm includes the following steps:

Use the Relative Deviation Index (RDI) to compare the target values of each combination optimization solution algorithm under each simulation example. The specific calculation formula is as follows:

Among them, RDI _IM represents the relative deviation value of the combined optimization solution algorithm M under the simulation example I; I represents the simulation example; FO _IM represents the simulation result of the combined optimization solution algorithm M under the simulation example I; Best _I and Worst _I Respectively represent the optimal results and the worst results of all combined optimization solution algorithms under simulation example I.

Preferably, in step S6, each of the simulation results is analyzed with different performance evaluation indicators. The performance evaluation indicators include the maximum completion time of the transportation task, the total transportation distance of the AGV when performing the transportation task, the total transportation distance of the AGV when performing the transportation task, The inbound transportation distance, the outbound transportation distance of AGV performing transportation tasks, cargo space utilization and AGV balance.

The technical solutions provided by the embodiments of this application may include the following beneficial effects:

In this solution, a composite algorithm framework is used to combine the cargo space allocation algorithm and the heavy-load AGV scheduling rule algorithm to form a variety of combined optimization solution algorithms, and simulation technology is used to simulate actual system state changes and production environment for scheduling. Among them, multiple simulation results obtained during the simulation process are analyzed with different performance evaluation indicators, so as to obtain the best combined optimization solution algorithm for each simulation result under different performance evaluation indicators. By applying the multi-load AGV storage and transportation strategy generated by the best-performing combinatorial optimization solution algorithm to the management of actual slate smart warehouses, the entry and exit of slate slabs can be completed in a timely manner, making the management of slate smart warehouses more efficient. Further enhance the market competitiveness of stone slab enterprises.

Description of drawings

Figure 1 is a step flow chart of a multi-load AGV storage and transportation method based on slate intelligent warehouse.

Detailed ways

Embodiments of the present invention are described in detail below. Examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements with the same or similar functions. The embodiments described below with reference to the drawings are exemplary and are only used to explain the present invention and are not to be construed as limitations of the present invention.

A multi-load AGV storage and transportation method based on slate intelligent warehouse includes the following steps:

Step S1: Construct a cargo space allocation algorithm. The cargo allocation algorithm includes a cargo space allocation algorithm related to the frequency of goods leaving the warehouse, a cargo space allocation algorithm based on the value of the goods, a random cargo space allocation algorithm, and the cargo space allocation closest to the connection point. Algorithm and cargo space allocation algorithm in intermediate storage areas;

Step S2: Construct a heavy-load AGV scheduling rule algorithm. The heavy-load AGV scheduling rule algorithm includes a first-come-first-serve rule algorithm, a nearest rule algorithm, a furthest rule algorithm, a highest idle rule algorithm, a shortest running distance rule algorithm, and a random rule algorithm. ;

Step S3: According to the cargo space allocation algorithm and the heavy-load AGV dispatching rule algorithm, generate multiple combination optimization solution algorithms through a composite algorithm framework;

Step S4: Build a simulation platform;

Step S5: Determine multiple simulation examples, use multiple combination optimization solving algorithms in the simulation platform to simulate and solve each of the simulation examples, and obtain multiple simulation results;

Step S6: Analyze different performance evaluation indicators for each of the simulation results to determine the best combined optimization solution algorithm for each of the simulation results under different performance evaluation indicators;

Step S7: Based on the best-performing combined optimization solution algorithm under different performance evaluation indicators of each of the simulation results, correspondingly generate a storage and transportation strategy for the multi-load AGV.

This solution is a multi-load AGV storage and transportation method based on slate intelligent warehouse, as shown in Figure 1. The first step is to construct a cargo space allocation algorithm. The cargo allocation algorithm includes cargo space allocation related to the frequency of goods leaving the warehouse. Algorithm, cargo space allocation algorithm based on the value of the goods, random cargo space allocation algorithm, cargo space allocation algorithm closest to the connection point and cargo space allocation algorithm in the intermediate storage area. In this embodiment, the multi-load AGV storage and transportation problem in the slate smart warehouse is decomposed into two sub-problems, namely the cargo space allocation sub-problem and the AGV scheduling sub-problem. For the sub-problem of cargo space allocation, five cargo space allocation algorithms were constructed by dividing the cargo space status and considering the frequency of goods leaving the warehouse, the value of the goods, the arrival time of the goods, the inbound transportation distance and the outbound transportation distance. When the goods arrive at the docking station, the goods are allocated using the cargo location allocation algorithm. During allocation, the status of all relevant current cargo locations needs to be known. The cargo location status represents the relationship between the cargo location and the product object. The cargo location The status is divided into free, occupied, picked, reserved and damaged. To further explain, the Location Allocation Algorithm for Associated Goods Outgoing Frequency (LAAAGOF for short) refers to the allocation of cargo locations based on the outgoing frequency of the goods. Goods with a higher warehouse frequency are allocated to cargo locations closer to the unloading point, while goods with a lower outbound frequency are allocated to cargo locations farther away from the unloading point. Location Allocation Algorithm Based on Commodity Value (LAABCV for short) means that when allocating goods, the value of the goods is taken into consideration, and goods with high value and frequent shipments are prioritized. In the cargo space closer to the unloading point. Random Location Allocation Algorithm (RLAA for short) refers to randomly selecting vacant storage areas when allocating cargo locations. The Algorithm for Assigning the Closest Storage Space to the Receiving Point (AACSSRP for short) means that when the goods arrive at the connection station, the frequency of shipment of various types of goods is considered. Frequency, the goods with high frequency of leaving the warehouse will be prioritized in the storage areas closer to the connecting station, and the goods with low frequency of leaving the warehouse will be prioritized in the storage areas far away from the connecting station. The Algorithm for the Allocation ofStorage Spaces in the Intermediate Storage Area (AASSISA) refers to the consideration of the outbound frequency of various types of goods when the goods arrive at the connection station. According to the outbound frequency of the goods, Goods with a high delivery rate are prioritized in the intermediate storage area, and goods with low delivery frequency are prioritized in storage areas far away from the docking station.

The second step is to construct an overloaded AGV scheduling rule algorithm. The overloaded AGV scheduling rule algorithm includes a first-come, first-served rule algorithm, a nearest rule algorithm, a furthest rule algorithm, a highest idle rule algorithm, a shortest running distance rule algorithm, and a random rule. algorithm. In this embodiment, for the AGV scheduling sub-problem, six AGV scheduling rule algorithms are constructed by considering AGV priority, distance between AGV and connection point, distance between AGV and unloading point, AGV idle rate, AGV transportation distance and random selection. . During the daytime production process, when the goods in the workshop are processed and arrive at the connection station, and the goods are allocated through the goods location allocation algorithm, the warehousing system calls the AGV scheduling rule algorithm to assign tasks to the idle AGVs. The scheduled AGVs Transport the goods from the docking station to the assigned cargo space; during the order outgoing process at night, when customer orders arrive, the warehousing system calls the AGV scheduling rule algorithm to assign tasks to idle AGVs. The scheduled AGVs are assigned tasks based on order information and warehousing The goods storage information in the area transports the goods from the target cargo location to the unloading point, and then returns to the designated stopping point. To further explain, the first-come-first-serve rule algorithm (FCFS for short) specifically includes the following steps: The first step is to initialize the idle car set Q, and add all idle cars to the idle car set Q in order of number; the second step is when the transportation task arrives. Then select the first element (AGV) in the set and remove it from the idle car set Q; otherwise wait for a new idle car to join the free car set Q, then select the first element in the set and remove it from the free car set Q Remove it from the car set Q; the third step is that after the called car completes the transportation task, it returns to the designated stopping point, the car status is set to the idle state, and is added to the |Q|th element in the idle car set Q . The nearest rule algorithm (referred to as Nearest) specifically includes the following steps: The first step is to initialize the idle car set Q, and add all idle cars to the idle car set Q in order of number; the second step is when the transportation task arrives, if/> Calculate the distance between each car in the idle car set Q and the transportation task, and obtain the distance set D of the cars. The element numbers are represented by 1, 2,...c; the third step is to find the car closest to the transportation task: d _c =Min{ d ₁ , d ₂ ,…d _c }, remove d _c from the car’s distance set D, call car c, and remove car c from the idle car set Q; Step 4 is that the called car completes transportation After the task, return to the designated stopping point, set the car status to the idle state, and join the idle car set Q, and update the car's distance set Q. The farthest rule algorithm (Farthest for short) specifically includes the following steps: Step 1 is to initialize the set of idle cars Q, and add all idle cars to the set of idle cars Q in order of number; Step 2 is when the transportation task arrives, if/> Calculate the distance between each car in the idle car set Q and the transportation task, and obtain the distance set Q of the cars. The element numbers are represented by 1, 2,...c; the third step is to find the car farthest from the transportation task: d _c = Max {d ₁ , d ₂ ,…d _c }, remove d _c from the car’s distance set D, call the car c, and remove the car c from the idle car set Q; Step 4 is the completion of the called car After the transportation task, return to the designated stopping point, set the car status to the idle state, and join the idle car set Q, and update the car's distance set D. The highest idle rule algorithm (HighestIdle for short) specifically includes the following steps:

The first step is to initialize the idle car set Q, and add all idle cars to the idle car set Q in order of number; the second step is when the transportation task arrives. If Calculate the idle rate of each car in the idle car set Q, that is, the car transportation time/the sum of the car transportation time and the car waiting time, and obtain the idle rate set F of the car. The element number is represented by 1, 2,...c; the third step is to find The car farthest from the transportation task: f _c = Max {f ₁ , f ₂ ,...f _c }, remove f _c from the car’s idle rate set F, call car c, and remove car c from the idle car Removed from the set Q; the fourth step is that after the called car completes the transportation task, it returns to the designated stopping point, the car status is set to the idle state, it is added to the idle car set Q, and the idle rate set F of the car is updated. Running the shortest distance rule algorithm (referred to as ShortestDistance) specifically includes the following steps:

The first step is to initialize the idle car set Q and the car driving distance set S. The element numbers are represented by 1, 2,...c; the second step is the arrival of the transportation task. If Calculate the running distance s _c of each car in the idle car set Q; the third step is to find the car with the shortest total travel distance: s _c = Min {s ₁ , s ₂ ,...s _c }, call car c, and add car c Removed from the idle car set Q; Step 4 is that after the called car completes the transportation task, it returns to the designated stopping point, the car status is set to the idle state, and is added to the idle car set Q, and the car travel distance set S is updated. The random rule algorithm (Random for short) specifically includes the following steps: The first step is to initialize the idle car set Q, and the element numbers are represented by 1, 2,...c; the second step is the arrival of the transportation task, if/> Randomly select a car: c=Rand({1,2,…c}), Rand is a random method, which is used to randomly select elements in the specified set, call the car c, and move the car c from the free car set Q Except; the third step is that after the called car completes the transportation task, it returns to the designated stopping point, the car status is set to the idle state, and it joins the idle car set Q.

The third step is to generate a variety of combined optimization solution algorithms through a composite algorithm framework based on the cargo space allocation algorithm and the heavy-load AGV dispatching rule algorithm. In this embodiment, 5 cargo location allocation algorithms and 6 AGV scheduling rule algorithms form 30 combinatorial optimization solution algorithms through a composite algorithm framework. The 30 combinatorial optimization solution algorithms are beneficial to subsequent use of them to solve simulation examples.

The fourth step is to build a simulation platform. In this embodiment, a Plant Simulation simulation platform is built based on the actual workshop layout, and the cargo space allocation algorithm and AGV scheduling rule algorithm are embedded.

The fifth step is to determine multiple simulation examples, and use a variety of combined optimization solving algorithms in the simulation platform to simulate and solve each of the simulation examples to obtain multiple simulation results. In this embodiment, combined with the production scenario of a stone slab enterprise, a simulation example is determined in which the influencing factors are AGV speed in no-load state, AGV speed in loaded state, connection station capacity, order arrival time interval and task type ratio. For each simulation example, 30 combination optimization solving algorithms are used for simulation solution, and the simulation results corresponding to each simulation example are obtained, which can test and analyze the effectiveness and suitability of the algorithm design scheme.

The sixth step is to analyze different performance evaluation indicators for each of the simulation results to determine the best combined optimization solution algorithm for each of the simulation results under different performance evaluation indicators. In this embodiment, the performance evaluation indicators include the maximum completion time of the transportation task, the total transportation distance of the AGV when performing the transportation task, the inbound transportation distance of the AGV when performing the transportation task, the outbound transportation distance of the AGV when performing the transportation task, and cargo space. Utilization rate and AGV balance. In order to determine the best algorithm among the proposed combinatorial optimization solving algorithms, the relative deviation index (RDI) is used to compare the target values of each combinatorial optimization solving algorithm under each simulation example, which is helpful for analyzing different impacts. The influence of factors on different performance evaluation indicators of target values.

The seventh step is to generate a storage and transportation strategy for multi-load AGV based on the best combined optimization solution algorithm under different performance evaluation indicators of each of the simulation results. Specifically, the best-performing combinatorial optimization solution algorithm enables the most reasonable cooperation between the cargo location allocation algorithm and the AGV scheduling rule algorithm. By applying the multi-load AGV storage and transportation strategy generated by the best-performing combinatorial optimization solution algorithm to the management of actual slate smart warehouses, the entry and exit of slate slabs can be completed in a timely manner, making the management of slate smart warehouses more efficient. Further enhance the market competitiveness of stone slab enterprises.

Preferably, in step S1, before constructing the cargo space allocation algorithm, the following steps are also included:

Step S11: Determine the cargo encoding method;

Step S12: Use the cargo coding method to code the cargo locations in the smart warehouse;

Step S13: Divide the status of the coded goods.

In this embodiment, the cargo location coding method is first determined, and then the unified cargo coding method is used to uniquely code the cargo locations in the smart warehouse in order to achieve dynamic binding of goods and cargo locations. Then, the cargo locations are dynamically bound according to the characteristics of the slate slabs in and out of the warehouse. The cargo location status is reasonably divided, and finally a cargo location allocation algorithm is constructed. It is further explained that the division of cargo space codes and cargo space status in the storage area is to facilitate the management and use of the cargo space resources in the storage area. The cargo location coding generally uses letters, numbers or symbols to represent the location, size and other information of the cargo location, so that the cargo location can be found quickly and accurately. The cargo space status is to classify the usage of the cargo space, such as free, occupied, picked, reserved and damaged, so as to understand the utilization of the cargo space. Cargo location allocation is closely related to cargo location coding and cargo location status division. When allocating cargo space, you need to consider the status of the cargo space and give priority to the available cargo space for allocation. At the same time, the occupied or reserved cargo space needs to be tracked and the occupied cargo space resources can be released in time to facilitate the storage of the next batch of goods.

Preferably, in step S1, in the cargo space allocation algorithm associated with the frequency of goods leaving the warehouse, the calculation formula for the frequency of goods leaving the warehouse is as follows:

Among them, GI _i represents the outbound frequency of goods of category i, i=1, 2,...I; GN _i represents the outbound quantity of goods of category i in the past period;

In the cargo location allocation algorithm based on the value of the goods, the value calculation formula of the goods is as follows:

Among them, GVI _i represents the value coefficient of goods of type i, i=1, 2,...I; G _i represents the specifications of goods of type i, expressed by the number of standard cargo spaces occupied by the goods; GN _i represents the number of standard goods of type i The quantity of goods shipped out of the warehouse in the past period.

In this embodiment, in the cargo location allocation algorithm associated with the frequency of goods leaving the warehouse, the frequency of goods leaving the warehouse is derived from past historical order data. When the arriving goods are not in the shipment frequency data of the goods, the frequency will be defaulted to 0 (new product). After the shipment of a batch of orders is completed, the shipment frequency of the goods will be updated. The greater the value of the goods' outbound frequency, the closer the goods will be to the unloading point. In the cargo location allocation algorithm based on the value of the goods, the greater the value coefficient of the goods, the closer the goods will be to the unloading point.

Preferably, in step S5, before determining multiple simulation examples, the following steps are also included:

Determine the influencing factors of the simulation example. The influencing factors of the simulation example include AGV speed in no-load state, AGV speed in loaded state, connection station capacity, order arrival time interval and task type ratio.

In this embodiment, determining the impact factors of the simulation example is helpful for subsequent analysis of its impact on different performance evaluation indicators of the target value. Further explanation, the AGV speed has two levels of 1.0m/s and 1.2m/s in the no-load state; the AGV speed has two levels of 0.6m/s and 0.8m/s in the loaded state; the connection station capacity is 1 or 2; There are three levels of order arrival time intervals, which are Poisson distributions with mean values of 5min, 10min, and 15min. Task types are divided into inbound tasks and outbound tasks. The ratio of task types refers to the inbound quantity and outbound quantity. Ratio has three levels: low, medium and high, with corresponding values 1/7, 3/7 and 5/7.

Preferably, in step S6, the determination of the best performing combinatorial optimization solution algorithm includes the following steps:

Use the Relative Deviation Index (RDI) to compare the target values of each combination optimization solution algorithm under each simulation example. The specific calculation formula is as follows:

Among them, FO _IM represents the relative deviation value of the combined optimization solution algorithm M under the simulation example I; I represents the simulation example; FO _IM represents the simulation result of the combined optimization solution algorithm M under the simulation example I; Best _I and Worst _I Respectively represent the optimal results and the worst results of all combined optimization solution algorithms under simulation example I.

In this embodiment, by using the relative deviation index to compare the target values of each combination optimization solution algorithm under each simulation example, it is helpful to determine the best algorithm among all combination optimization solution algorithms under each simulation example, and further The multi-load AGV storage and transportation strategy generated by this optimal algorithm is applied to the management of actual slate smart warehouses.

Preferably, in step S6, each of the simulation results is analyzed with different performance evaluation indicators. The performance evaluation indicators include the maximum completion time of the transportation task, the total transportation distance of the AGV when performing the transportation task, the AGV when performing the transportation task The inbound transportation distance, the outbound transportation distance of AGV performing transportation tasks, cargo space utilization and AGV balance.

In this embodiment, when each of the simulation results is analyzed using the maximum completion time of the transportation task as the performance evaluation index, the best algorithm among the 30 combination optimization solving algorithms is the RLAA-Nearest combination algorithm. When analyzing each of the simulation results using the total transportation distance of the AGV performing transportation tasks as the performance evaluation index, the cargo space allocation algorithm LAABCV performed better overall, and can be combined with the AGV scheduling rule algorithms Nearest and Random to obtain smaller results. Total transportation distance. When each of the simulation results was analyzed using the warehousing transportation distance of the AGV performing transportation tasks as the performance evaluation index, the best algorithm among the 30 combination optimization solving algorithms was the AACSSRP_Farthest combination algorithm. When each of the simulation results was analyzed using the outbound transportation distance of the AGV performing transportation tasks as the performance evaluation index, the best algorithm among the 30 combination optimization solving algorithms was the LAAAGOF_FCFS combination algorithm. When analyzing each of the simulation results based on the performance evaluation index of cargo space utilization, since the cargo space utilization in the storage area does not change much under different AGV scheduling rule algorithms, only the best performance among different cargo space allocation algorithms is considered. The best algorithm, the cargo space allocation algorithm AACSSRP is used in this embodiment, which can make the average utilization rate of cargo spaces in each storage area of the rock slab warehouse reach more than 70%. When each of the simulation results was analyzed using AGV balance as the performance evaluation index, the best algorithm among the 30 combination optimization solving algorithms was the AACSSRP-ShortestDistance combination algorithm.

In addition, each functional unit in various embodiments of the present invention can be integrated into a processing module, or each unit can exist physically alone, or two or more units can be integrated into one module. The above integrated modules can be implemented in the form of hardware or software function modules. If the integrated module is implemented in the form of a software function module and sold or used as an independent product, it can also be stored in a computer-readable storage medium.

Although the embodiments of the present invention have been shown and described above, it can be understood that the above-mentioned embodiments are illustrative and should not be construed as limitations of the present invention. Those of ordinary skill in the art can make modifications to the above-mentioned embodiments within the scope of the present invention. The embodiments are subject to changes, modifications, substitutions and variations.

Claims (6)

1. A multi-load AGV storage and transportation method based on slate intelligent warehouse, which is characterized by: including the following steps: Step S1: Construct a cargo space allocation algorithm. The cargo allocation algorithm includes a cargo space allocation algorithm related to the frequency of goods leaving the warehouse, a cargo space allocation algorithm based on the value of the goods, a random cargo space allocation algorithm, and the cargo space allocation closest to the connection point. Algorithm and cargo space allocation algorithm in intermediate storage areas; Step S2: Construct a heavy-load AGV scheduling rule algorithm. The heavy-load AGV scheduling rule algorithm includes a first-come-first-serve rule algorithm, a nearest rule algorithm, a furthest rule algorithm, a highest idle rule algorithm, a shortest running distance rule algorithm, and a random rule algorithm. ; Step S3: Based on the cargo space allocation algorithm and the heavy-load AGV dispatching rule algorithm, generate 30 combination optimization solution algorithms through a composite algorithm framework; Step S4: Build a simulation platform; Step S5: Determine multiple simulation examples, use multiple combination optimization solving algorithms in the simulation platform to simulate and solve each of the simulation examples, and obtain multiple simulation results; Step S6: Analyze different performance evaluation indicators for each of the simulation results to determine the best combined optimization solution algorithm for each of the simulation results under different performance evaluation indicators; Step S7: Based on the best-performing combined optimization solution algorithm under different performance evaluation indicators of each of the simulation results, correspondingly generate a storage and transportation strategy for the multi-load AGV. 2. A multi-load AGV storage and transportation method based on slate intelligent warehouse according to claim 1, characterized in that: in step S1, before constructing the cargo space allocation algorithm, the following steps are also included: Step S11: Determine the cargo encoding method; Step S12: Use the cargo coding method to code the cargo locations in the smart warehouse; Step S13: Divide the status of the coded goods. 3. A multi-load AGV storage and transportation method based on slate intelligent warehouse according to claim 1, characterized in that: in step S1, in the cargo space allocation algorithm related to the frequency of goods leaving the warehouse, the number of goods is The formula for calculating the outbound frequency is as follows: Among them, GI _i represents the outbound frequency of goods of category i, i=1, 2,...I; GN _i represents the outbound quantity of goods of category i in the past period; In the cargo location allocation algorithm based on the value of the goods, the value calculation formula of the goods is as follows: Among them, GVI _i represents the value coefficient of goods of type i, i=1, 2,...I; G _i represents the specifications of goods of type i, expressed by the number of standard cargo spaces occupied by the goods; GN _i represents the number of standard goods of type i The quantity of goods shipped out of the warehouse in the past period. 4. A multi-load AGV storage and transportation method based on slate intelligent warehouse according to claim 1, characterized in that: in step S5, before determining multiple simulation examples, the following steps are also included: Determine the influencing factors of the simulation example. The influencing factors of the simulation example include AGV speed in no-load state, AGV speed in loaded state, connection station capacity, order arrival time interval and task type ratio. 5. A multi-load AGV storage and transportation method based on slate intelligent warehouse according to claim 1, characterized in that: in step S6, the determination of the best performing combination optimization solution algorithm includes the following steps: Use the Relative Deviation Index (RDI) to compare the target values of each combination optimization solution algorithm under each simulation example. The specific calculation formula is as follows: Among them, RDI _IM represents the relative deviation value of the combined optimization solution algorithm M under the simulation example I; I represents the simulation example; FO _IM represents the simulation result of the combined optimization solution algorithm M under the simulation example I; Best _I and Worst _I Respectively represent the optimal results and the worst results of all combined optimization solution algorithms under simulation example I. 6. A multi-load AGV storage and transportation method based on slate intelligent warehouse according to claim 1, characterized in that: in step S6, different performance evaluation indicators are analyzed for each of the simulation results. Performance evaluation indicators include the maximum completion time of the transportation task, the total transportation distance of the AGV when performing the transportation task, the inbound transportation distance of the AGV when performing the transportation task, the outbound transportation distance of the AGV when performing the transportation task, cargo space utilization and AGV balance. sex.