Order goods sorting method based on block center of gravity greedy idea

Abstract

The present invention's method for sorting ordered goods based on a block-based, center-of-gravity greedy approach involves first establishing a waiting area and an order area, assuming there are y pallets in the waiting area and x pallets in the order area. The method then calculates the score of goods on pallet k in the waiting area containing goods with order number b being sorted onto pallet a in the order area, obtaining a set B of scores for goods with all order numbers being sorted onto each pallet in the order area. The order number j with the highest score is then matched with pallet i in the order area, and goods are sorted according to the established correspondence between order numbers and pallets in the order area. Compared to existing methods using random location distribution or genetic algorithms, the present invention's method for sorting ordered goods reduces overall sorting time and achieves higher sorting efficiency, making it suitable for applications requiring high sorting efficiency.

Description

Technical Field

[0001] The present invention relates to a cargo sorting method, and more specifically, to an order cargo sorting method based on a greedy block gravity center idea and applied to an aerial sorting truss manipulator. Background Art

[0002] In logistics companies, manufacturing companies’ workshops and other places, large and bulky goods need to be sorted, which are usually grasped and sorted by aerial sorting truss robots. Figure 1 As shown, a structural schematic diagram of an existing aerial sorting truss manipulator device is given, which consists of a frame, an elevated track, a bridge (commonly known as a trolley), a trolley and a grasping device. The bridge shown is laid on two parallel elevated tracks and can move along the length direction of the elevated tracks; the trolley is set on the bridge and can run along its length direction, and the grasping terminal set on the trolley is a grasping device such as a manipulator or an industrial robot; therefore, the aerial sorting truss manipulator can be adapted to applications under various working conditions, such as the lifting, transportation and classification of goods in workshops, warehouses and other occasions of manufacturing enterprises such as servers, tires, floors and building materials.

[0003] like Figure 2 As shown in the figure, the sorting operation flow chart of the existing aerial sorting truss manipulator device is given. The total sorting time of the aerial sorting truss manipulator system includes the horizontal movement time and the vertical lifting, grasping and placing time. The operation process is as follows Figure 2 As shown in the figure, for different work tasks, since the lifting height of the grasping device is the same, the speed and acceleration of the movement along the Z-axis (vertical direction) are the same, and the time for executing a single grasping and placing task is the same, the vertical lifting, grasping and placing time are fixed values. The main difference in the operation time of different tasks is the horizontal movement time.

[0004] The horizontal movement time includes the movement time when the system is without goods, namely the "empty movement time", and the movement time when the system is with goods, namely the "loaded movement time". The "empty movement time" can adopt the "nearest pick-up" operation mode. Therefore, how to optimize the "loaded movement time" has become the key to speeding up the efficiency of cargo sorting. This paper mainly improves the efficiency of cargo sorting by minimizing the "loaded movement time" after the aerial sorting truss robot grabs the goods. Summary of the Invention

[0005] In order to overcome the shortcomings of the above technical problems, the present invention provides an order goods sorting method based on the greedy idea of ​​block center of gravity.

[0006] The method for sorting order goods based on the block center of gravity greedy idea of ​​the present invention is characterized by being implemented by the following method: first, establishing a waiting area and an order area, wherein the waiting area is used to place pallets containing goods to be sorted, and the order area is used to place pallets for storing goods with order numbers. It is assumed that there are y pallets in the waiting area and x pallets in the order area; then, the order numbers and corresponding quantities of the goods on each pallet in the waiting area are counted, and the score of the corresponding goods on pallet k in the waiting area containing goods with order number b being sorted to pallet a in the order area is calculated using formula (1):

[0007]

[0008] In formula (1), 1≤b≤m, m is the total order quantity, and m≤x; 1≤k≤y, 1≤a≤x;

[0009] Formula (1) is used to calculate the scores of the corresponding goods on each pallet containing goods with order number b in the picking area being sorted to pallet a in the order area, and the obtained scores are summed to obtain the scores of the goods with order number b being sorted to pallet a in the order area; similarly, the scores of the pallets containing goods with order number b in the picking area being assigned to each pallet in the order area are calculated; using the same calculation method, the score set B of the goods with all order numbers being sorted to each pallet in the order area is obtained;

[0010] In set B, a greedy approach is adopted to match the order number j with the highest score with the pallet i in the order area. That is, pallet i in the order area is used as the pallet for storing the goods with order number j, and a matching correspondence between order number j and pallet i is established. The matched order number j and pallet i in the order area are deleted from the orders and pallets in the order area, and the matching of order numbers and pallets in the order area is continued until all order numbers are matched. After the matching is completed, the goods are sorted according to the established correspondence between order numbers and pallets in the order area.

[0011] The order goods sorting method based on the block center of gravity greedy idea of ​​the present invention is specifically implemented by the following steps:

[0012] a) Area division: The central area below the aerial truss robot device is divided into a waiting area, and the two sides of the waiting area are divided into order areas. The waiting area is where pallets storing goods to be sorted are placed, and the order area is where pallets storing goods with order numbers are placed;

[0013] b) Cargo coding. Logistics vehicles transport mixed stacked pallets to the sorting area. Workers scan the identification code of the goods on each pallet to be sorted, record the cargo information on each pallet, and push the scanned pallets into the picking area in sequence. Then, empty pallets for storing ordered goods are placed in the order area. The number of pallets in the picking area is y, and their labels are Z1, Z2, .., Z. Y , the number of pallets in the order area is x, and their labels are S1, S2, .., S x ;

[0014] c) Obtain order number information, order quantity, and item information for each order number. Suppose the order quantity is m, and a pallet in the order area is used to store all items belonging to order number j, where m ≤ x.

[0015] d) Obtain the quantity matrix of the goods with the order number, and determine the number of all current order numbers on the pallets Z1 to Z1 in the picking area according to the goods identification code on the pallet to be sorted recorded by scanning in step b). y The goods quantity matrix C on:

[0016]

[0017] In the matrix C, Z ij Indicates pallet Z in the picking area i The number of goods with order number j, Z 11 , Z 21 ,...,Z y1 Respectively represent order number 1 in the picking area tray Z1 to Z y The quantity of goods, and so on;

[0018] e) Establish a distance inverse value matrix. By calculating the distance between each pallet in the picking area and each pallet in the order area, establish the distance inverse value matrix D between the pallets in the picking area and the pallets in the order area:

[0019]

[0020] In the matrix D, 1 / (L zisj ) indicates the pallet Z in the picking area i With the pallet S in the order area j The reciprocal of the distance between z1s1 )、1 / (L z2s1 ),...,1 / (L zys1 ) represent the pallets Z1 to Z2 in the picking area y The reciprocal of the distance from the order area pallet S1, and so on;

[0021] f) Obtain a scoring matrix by multiplying the goods quantity matrix C obtained in step d) by the distance reciprocal matrix D obtained in step e) to obtain the scoring matrix E for all goods with all order numbers on all pallets in the order area:

[0022]

[0023] In the matrix E, S ij Indicates that the goods in order number j are sorted to pallet S in the order area i Rating when last used;

[0024] g). Match the order number and the pallet, find the maximum value of the elements in the matrix E, and let the element with the maximum value in the matrix E be S ij , then the order number j is associated with the pallet S in the order area i Match them together and use order number j with pallet S i The final matching result is stored; at the same time, the matched order number j is deleted from the order number, and the matched pallet S is i Remove from order area tray;

[0025] h) Determine whether all order numbers have been matched. If so, proceed to step i); if not, proceed to steps d) to g) to match the next order number with the pallet in the order area.

[0026] i). Goods sorting: After all order numbers are matched, goods are sorted according to the established correspondence between all order numbers and pallets in the order area.

[0027] In the method for sorting order goods based on the greedy idea of ​​block center of gravity of the present invention, in steps d) and e), if the pallet S in the order area j Since it has established a matching relationship with order number i and is deleted from the order area pallet and order number, the order number i in the picking area pallet Z1 to Z is no longer calculated in step d). y The number of goods on the order area is not calculated in step e) for each pallet in the picking area and the pallet S in the order area. j The distance between them.

[0028] In the method for sorting order goods based on the greedy idea of ​​block center of gravity of the present invention, the goods sorting in step i) is specifically achieved by the following steps:

[0029] i-1) First, the aerial sorting truss robot identifies the pallet of goods to be sorted that is closest to its current stop in the picking area. It then drives the large and small carts to move the grabbing terminal above the nearest pallet.

[0030] i-2). Then, scan the barcode on the outer packaging of the goods to identify the goods, and use the grabbing terminal to grab and lift the goods after the barcode is identified;

[0031] i-3) Next, based on the matching relationship between the order number and the pallet in the order area established in steps g) and h), the pallet number in the order area where the currently grabbed item should be placed is determined. The large and small carts are then driven to move the item to the pallet where it should be placed. The grabbing terminal releases the item and places it on the pallet in the corresponding order area, completing a sorting operation.

[0032] i-4) After a sorting operation is completed, the aerial sorting truss robot searches for the nearest pallet in the picking area based on its current position and performs the next cargo sorting according to steps i-2) to i-3) until all the cargo in the picking area is sorted to the order area and the sorting task is completed.

[0033] The beneficial effects of the present invention are as follows: the order goods sorting method based on the block gravity center greedy idea of ​​the present invention first establishes a waiting area and an order area under the aerial sorting truss manipulator device, and divides the blocks according to the principle that each pallet in the waiting area and the order area is a block, and divides the waiting area into y blocks corresponding to the number and position of the pallets therein, and divides the order area into x blocks corresponding to the number and position of the pallets therein, making use of the "blocking" idea; then, based on the principle that "the number of goods in the order number on the pallet in the waiting area is equal to the number of goods in the order area" The method uses the "ratio of distances between pallets" as a scoring calculation method to calculate the score of all goods in each order sorted to each pallet in the order area, and then matches the order number with the highest score with the pallet in the sorting area. The principle is to move goods with the same order number on the pallet in the picking area to the pallet in the order area that is closer to its center of gravity, utilizing the "center of gravity" concept. After each order number is matched with the pallet in the picking area based on the maximum score, it is deleted from the order and the pallet in the picking area, and then re-matched, which is characterized by a "greedy" concept. Therefore, compared with existing random distribution methods and genetic algorithms for cargo locations, the order goods sorting method of the present invention based on the center of gravity greedy concept reduces the overall time required for sorting goods and improves the sorting efficiency, making it suitable for applications with high sorting efficiency requirements. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Figure 1 This is a structural diagram of an existing aerial sorting truss manipulator device;

[0035] Figure 2 This is a sorting operation flow chart of the existing aerial sorting truss manipulator device;

[0036] Figure 3A diagram showing the area below the aerial sorting truss robot device of the present invention;

[0037] Figure 4 Schematic diagram of the aerial sorting truss manipulator device of the present invention sorting goods for the i-th time;

[0038] Figure 5 This is a schematic diagram of the cargo sorting process in the present invention;

[0039] Figure 6 Schematic diagram of the principle process of the order goods sorting method based on the center-of-gravity greedy idea of ​​the present invention;

[0040] Figure 7 This is the relationship curve between the average load movement time and different order dispersion obtained in the control experiment.

[0041] In the picture: 1 frame, 2 elevated track, 3 bridge, 4 trolley, 5 grab terminal. DETAILED DESCRIPTION

[0042] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0043] like Figure 1 Figure 1 shows a schematic diagram of the structure of a conventional aerial sorting truss robot device. It consists of a frame 1, elevated tracks 2, a bridge 3, a trolley 4, and a gripping terminal 5. The frame 1 serves as a fixture and support. Two elevated tracks 2 are fixed parallel to the frame 1. The ends of the bridge 3 (also known as a trolley) are mounted on the guide rails of the elevated tracks 2, allowing the bridge 3 to move along the length of the elevated tracks 2. In the established rectangular coordinate system XYZ, the elevated tracks 2 align with the X-axis, and the bridge 3 aligns with the Y-axis. The trolley 4 is mounted on the bridge 3 and can move along the length of the bridge 3. The gripping terminal 5 is located at the lower end of the trolley 4. The gripping terminal 5 can utilize a pneumatic or electric hand gripper, or a gripping device such as an industrial robot. Therefore, the aerial sorting truss robot device is suitable for a variety of applications, such as the lifting, transportation, and sorting of goods in workshops and warehouses of manufacturing enterprises producing servers, tires, flooring, and building materials.

[0044] The finished product warehouse of manufacturing enterprises uses aerial sorting truss manipulators to complete the sorting of mixed pallet finished products before storage. The lower part of the aerial sorting truss manipulator device is divided into the waiting area and the order area, such as Figure 3The figure shows the area division beneath the aerial sorting truss robot device in the present invention. For ease of study, the pallets in the to-be-picked area and the order area are numbered, with a total of y locations for the to-be-picked area and x locations for the order area. The to-be-picked area is located in the center of the system and temporarily stores mixed pallets of finished goods coming off the production line. The order areas are located on either side of the to-be-picked area and temporarily store pallets of ordered goods. In-plant logistics vehicles deliver mixed pallets of goods coming off the production line to the finished goods warehouse. On-site workers move all pallets of goods from the vehicle compartment to the to-be-picked area for storage. The stacking locations of the goods on each pallet are determined by manually scanning the box barcodes in a fixed sequence. A database of the stacking locations of the goods in the to-be-picked area is then established by the aerial sorting truss robot, which then sorts and stacks the goods in the to-be-picked area onto pallets in the order area according to the shipping order. The aerial sorting truss robot's operating area covers both the to-be-picked area and the order area.

[0045] like Figure 2 As shown in the figure, the sorting operation flow chart of the existing aerial sorting truss manipulator device is given. The total sorting time of the aerial sorting truss manipulator device includes the horizontal movement time and the vertical lifting, grasping and placing time. The operation process is as follows Figure 2 As shown in the figure, for different tasks, the grasping device has the same lifting height, the same speed and acceleration along the Z axis, and the same time required to execute a single grasping and placing task. Therefore, the vertical grasping and placing time is fixed, and the main difference in the operation time of different tasks is the horizontal movement time.

[0046] The horizontal movement time includes the movement time when the system has no goods, that is, the "no-load movement time", and the movement time when the system has goods, that is, the "load movement time". Figure 4 As shown in the figure, a schematic diagram of the i-th cargo sorting by the aerial sorting truss manipulator device of the present invention is given. After determining the number of work tasks, the vertical grabbing and placing time of the system is fixed; the "empty load movement time" adopts the "nearest pick-up" operation mode, that is, after the current goods are placed, the location information of the nearest waiting area with goods is determined according to the location of the aerial sorting truss manipulator trolley, and the position is then moved to perform the lifting operation; the "cargo location allocation algorithm" is designed to optimize the "load movement time" of the truss sorting robot.

[0047] like Figure 6 As shown in FIG, a schematic diagram of the principle process of the order goods sorting method based on the block center of gravity greedy idea of ​​the present invention is given. The core idea of ​​the algorithm design of the present invention is as follows:

[0048] First, considering the number of goods and the distance between the smallest block in the picking area and the pallet in the order area, the score of the corresponding goods on pallet k in the picking area containing the goods in order number b being sorted to pallet a in the order area is calculated using formula (1):

[0049]

[0050] In formula (1), 1≤b≤m, m is the total order quantity, and m≤x; 1≤k≤y, 1≤a≤x;

[0051] Formula (1) is used to calculate the scores of the corresponding goods on each pallet containing goods with order number b in the picking area being sorted to pallet a in the order area, and the obtained scores are summed to obtain the scores of the goods with order number b being sorted to pallet a in the order area; similarly, the scores of the pallets containing goods with order number b in the picking area being assigned to each pallet in the order area are calculated; using the same calculation method, the score set B of the goods with all order numbers being sorted to each pallet in the order area is obtained;

[0052] In set B, a greedy approach is adopted to match the order number j with the highest score with the pallet i in the order area. That is, pallet i in the order area is used as the pallet for storing the goods with order number j, and a matching correspondence between order number j and pallet i is established. The matched order number j and pallet i in the order area are deleted from the orders and pallets in the order area, and the matching of order numbers and pallets in the order area is continued until all order numbers are matched. After the matching is completed, the goods are sorted according to the established correspondence between order numbers and pallets in the order area.

[0053] The order goods sorting method based on the block center of gravity greedy idea of ​​the present invention is specifically implemented by the following steps:

[0054] a) Area division: The central area below the aerial truss robot device is divided into a waiting area, and the two sides of the waiting area are divided into order areas. The waiting area is where pallets storing goods to be sorted are placed, and the order area is where pallets storing goods with order numbers are placed;

[0055] b) Cargo coding. Logistics vehicles transport mixed stacked pallets to the sorting area. Workers scan the identification code of the goods on each pallet to be sorted, record the cargo information on each pallet, and push the scanned pallets into the picking area in sequence. Then, empty pallets for storing ordered goods are placed in the order area. The number of pallets in the picking area is y, and their labels are Z1, Z2, .., Z. Y , the number of pallets in the order area is x, and their labels are S1, S2, .., S x ;

[0056] c) Obtain order number information, order quantity, and item information for each order number. Suppose the order quantity is m, and a pallet in the order area is used to store all items belonging to order number j, where m ≤ x.

[0057] d) Obtain the quantity matrix of the goods with the order number, and determine the number of all current order numbers on the pallets Z1 to Z1 in the picking area according to the goods identification code on the pallet to be sorted recorded by scanning in step b). y The goods quantity matrix C on:

[0058]

[0059] In the matrix C, Z ij Indicates pallet Z in the picking area i The number of goods with order number j, Z 11 , Z 21 ,...,Z y1 Respectively represent order number 1 in the picking area tray Z1 to Z y The quantity of goods, and so on;

[0060] e) Establish a distance inverse value matrix. By calculating the distance between each pallet in the picking area and each pallet in the order area, establish the distance inverse value matrix D between the pallets in the picking area and the pallets in the order area:

[0061]

[0062] In the matrix D, 1 / (L zisj ) indicates the pallet Z in the picking area i With the pallet S in the order area j The reciprocal of the distance between z1s1 )、1 / (L z2s1 ),...,1 / (L zys1 ) represent the pallets Z1 to Z2 in the picking area y The reciprocal of the distance from the order area pallet S1, and so on;

[0063] Wherein, in steps d) and e), if the pallet S in the order area j Since it has established a matching relationship with order number i and is deleted from the order area pallet and order number, the order number i in the picking area pallet Z1 to Z is no longer calculated in step d). y The number of goods on the order area is not calculated in step e) for each pallet in the picking area and the pallet S in the order area. j The distance between them.

[0064] f) Obtain a scoring matrix by multiplying the goods quantity matrix C obtained in step d) by the distance reciprocal matrix D obtained in step e) to obtain the scoring matrix E for all goods with all order numbers on all pallets in the order area:

[0065]

[0066] In the matrix E, S ij Indicates that the goods in order number j are sorted to pallet S in the order area i Rating when last used;

[0067] It can be seen that the multiplication of the elements in matrix C and the elements in matrix D is equivalent to formula (1).

[0068] g). Match the order number and the pallet, find the maximum value of the elements in the matrix E, and let the element with the maximum value in the matrix E be S ij , then the order number j is associated with the pallet S in the order area i Match them together and use order number j with pallet S i The final matching result is stored; at the same time, the matched order number j is deleted from the order number, and the matched pallet S is i Remove from order area tray;

[0069] h) Determine whether all order numbers have been matched. If so, proceed to step i); if not, proceed to steps d) to g) to match the next order number with the pallet in the order area.

[0070] i). Goods sorting: After all order numbers are matched, goods are sorted according to the established correspondence between all order numbers and pallets in the order area.

[0071] like Figure 5 As shown in FIG, a schematic diagram of the cargo sorting process in the present invention is provided, and the cargo sorting described in step i) is specifically implemented by the following steps:

[0072] i-1) First, the aerial sorting truss robot identifies the pallet of goods to be sorted that is closest to its current stop in the picking area. It then drives the large and small carts to move the grabbing terminal above the nearest pallet.

[0073] i-2). Then, scan the barcode on the outer packaging of the goods to identify the goods, and use the grabbing terminal to grab and lift the goods after the barcode is identified;

[0074] i-3) Next, based on the matching relationship between the order number and the pallet in the order area established in steps g) and h), the pallet number in the order area where the currently grabbed item should be placed is determined. The large and small carts are then driven to move the item to the pallet where it should be placed. The grabbing terminal releases the item and places it on the pallet in the corresponding order area, completing a sorting operation.

[0075] After the first sorting operation, the aerial sorting truss manipulator searches for the nearest tray in the picking area based on its current position, and performs the next cargo sorting according to steps i-2) to step i-3) until all the goods in the picking area are sorted into the order area, and the sorting task is completed.

[0076] To verify the optimization effect of the order cargo sorting method based on the idea of greedy centroid of blocks on the operation scenario of the aerial sorting truss manipulator, the "random cargo location distribution method" is used as the control group, and the cargo location allocation scheme of the control group is randomly generated and disordered. The control group and the experimental group (the order cargo sorting method of the present invention) only use the cargo location allocation scheme as the variable, and other variables such as the order and the inbound sequence of goods are kept consistent. The orders adopted are all randomly generated by a computer program according to the actual situation on site, with the dispersion degree as the distinction of different groups of orders, and each dispersion degree contains multiple groups of orders.

[0077] In the order setting, 15 servers are placed on each inbound tray (regarded as full load), and 30 trays are inbound (filling the picking area), with a total of 450 goods to be sorted; the server production order numbers are set to 40 (the maximum number of server production order numbers that can be carried under the scenario layout in this article, corresponding to 40 trays (or cargo locations) in the order area, and the types of server production order numbers in the order can be less than 40).

[0078] Define the dispersion degree:

[0079] In the order, there are G loaded trays (in this case, G takes the value of 30), and the total number of goods on the gth (g ∈ G) tray is denoted as C (in this case, C takes the value of 15), and there are F (0 < F ≤ C) goods with production order numbers, and the number of goods belonging to the fth (f ∈ F) production order number is denoted as C f (f ∈ F, 0 ≤ C f ≤ C), then the dispersion degree of the goods on the gth tray (denoted as A g ) is given by the formula:

[0080]

[0081] where g ∈ G and G = 30, C = 15, 0 < F ≤ C and f ∈ F, 0 ≤ C f ≤ C.

[0082] If the overall dispersion degree of the G trays in the order (denoted as A) is to be calculated, then the sum of the dispersion degrees of each tray is calculated and the average value is taken, that is:

[0083]

[0084] where g ∈ G and G = 30, C = 15, 0 < F ≤ C and f ∈ F, 0 ≤ C f ≤ C.

[0085] The dispersion of goods on a single pallet, as defined by formula (5), is as follows: First, it is assumed that the dispersion is minimum when all goods on a pallet belong to the same production order number. Formula (5) reflects the overall deviation of goods with each production order number on the pallet from the optimal situation (i.e., all goods on the pallet belong to the same production order number). The smaller the calculated value, the smaller the overall deviation and the smaller the dispersion. The dispersion calculated above ranges from 0 to 196. For 40 orders, the load movement time results obtained using the "random distribution method" and the "block-based center-of-gravity greedy order goods sorting method of the present invention" are shown in Table 1.

[0086] Table 1

[0087]

[0088]

[0089] The no-load moving time of the aerial sorting truss manipulator is added to the loaded moving time in Table 1, and the total overall moving time results are shown in Table 2.

[0090] Table 2

[0091]

[0092]

[0093] The results of the comparison between the order goods sorting method based on the center-of-gravity greedy idea of ​​the present invention and the genetic algorithm are shown in Table 3:

[0094] Table 3

[0095]

[0096] It can be seen from Table 1 and Table 2 that the block center of gravity greedy cargo location allocation algorithm of the present invention will produce an optimization effect ranging from 14.7% to 45.4% compared with the load movement time of the cargo location random distribution method, with an average optimization rate of about 28.14%. The load movement time of the aerial sorting truss manipulator optimized by the method of this paper is significantly reduced compared with the cargo location random distribution method. The empirical results show that the cargo location allocation algorithm based on the block center of gravity idea proposed in this paper can effectively improve the load movement efficiency of the equipment. In addition, by Figure 7The data scatter plots and curves shown provide a visual representation of the performance of each algorithm. The scatter plot shows that for the block-based, center-of-gravity greedy algorithm, the scatter plots for different groups are relatively concentrated under the same order dispersion. This means that under similar order dispersions, the differences in load movement times are small. By studying the curve trends in the line graph and analyzing the dispersion, we find that when there is a certain gap in the dispersion of different orders, the load movement time of the aerial sorting truss robot obtained using the block-based, center-of-gravity greedy location allocation algorithm increases with increasing dispersion and tends to approach the load movement time of the random location distribution method. This indicates that as the dispersion of the inbound pallet servers increases, the effectiveness of the block-based, center-of-gravity greedy location allocation algorithm becomes increasingly limited. When the dispersion of the pallet servers remains at a relatively low level, the algorithm achieves a relatively good optimization effect.

[0097] Regarding the overall movement time of the aerial sorting truss robot, the block-based, center-of-gravity, greedy location allocation algorithm achieves optimization results ranging from 10.70% to 30.90% compared to the random location distribution method, with an average optimization rate of approximately 19.74%. While this rate is lower than the average optimization rate for load movement time alone, it still produces a relatively good optimization effect on the aerial sorting truss robot's operating time. Furthermore, the optimization rate of the block-based, center-of-gravity, greedy location allocation algorithm compared to the random location distribution method decreases with increasing dispersion. Therefore, the algorithm designed in this paper also experiences a decrease in optimization effect with increasing dispersion in overall movement time.

[0098] As can be seen from Table 3, the genetic algorithm produces an optimization effect of 0.4%-2.62% compared with the algorithm in this paper, with an average optimization rate of about 1.45%. The degree of optimization is low, and its CPU execution time is several thousand times that of the algorithm in this paper. The timeliness of field application is poor. Therefore, in summary, the block center of gravity greedy idea cargo allocation algorithm can achieve near-optimal operating efficiency while increasing the possibility of the algorithm being applied in real-time scenarios.

Claims

1. A method for sorting order goods based on the idea of ​​block center of gravity greedy is implemented by the following method: First, a waiting area and an order area are established. The waiting area is used to place pallets containing goods to be sorted, and the order area is used to place pallets for storing goods with order numbers. Suppose there are y pallets in the waiting area and x pallets in the order area. Then, the order numbers and corresponding quantities of goods on each pallet in the waiting area are counted, and the score of the corresponding goods on pallet k in the waiting area containing goods with order number b being sorted to pallet a in the order area is calculated using formula (1): In formula (1), 1≤b≤m, m is the total order quantity, and m≤x; 1≤k≤y, 1≤a≤x; Formula (1) is used to calculate the scores of the corresponding goods on each pallet containing goods with order number b in the picking area being sorted to pallet a in the order area, and the obtained scores are summed to obtain the scores of the goods with order number b being sorted to pallet a in the order area; similarly, the scores of the pallets containing goods with order number b in the picking area being assigned to each pallet in the order area are calculated; using the same calculation method, the score set B of the goods with all order numbers being sorted to each pallet in the order area is obtained; In set B, a greedy approach is used to match order number j with the highest score and pallet i in the order area. That is, pallet i in the order area is used as the pallet for storing goods with order number j, and a matching relationship between order number j and pallet i is established. The matched order number j and pallet i in the order area are deleted from the orders and pallets in the order area, and the matching process continues until all order numbers are matched. After the matching is completed, the goods are sorted according to the established correspondence between order numbers and pallets in the order area. It is characterized by: The order goods sorting method based on the block center greedy idea is implemented through the following steps: a) Area division: The central area below the aerial truss robot device is divided into a waiting area, and the two sides of the waiting area are divided into order areas. The waiting area is where pallets storing goods to be sorted are placed, and the order area is where pallets storing goods with order numbers are placed; b) Cargo coding. Logistics vehicles transport mixed stacked pallets to the sorting area. Workers scan the identification code of the goods on each pallet to be sorted, record the cargo information on each pallet, and push the scanned pallets into the picking area in sequence. Then, empty pallets for storing ordered goods are placed in the order area. The number of pallets in the picking area is y, and their labels are Z1, Z2, .., Z. Y , the number of pallets in the order area is x, and their labels are S1, S2, .., S x ; c) Obtain order number information, order quantity, and item information for each order number. Suppose the order quantity is m, and a pallet in the order area is used to store all items belonging to order number j, where m ≤ x. d) Obtain the quantity matrix of the goods with the order number, and determine the number of all current order numbers on the pallets Z1 to Z1 in the picking area according to the goods identification code on the pallet to be sorted recorded by scanning in step b). y The goods quantity matrix C on: In the matrix C, Z ij Indicates pallet Z in the picking area i The number of goods with order number j, Z 11 , Z 21 ,...,Z y1 Respectively represent order number 1 in the picking area tray Z1 to Z y The quantity of goods, and so on; e) Establish a distance inverse value matrix. By calculating the distance between each pallet in the picking area and each pallet in the order area, establish the distance inverse value matrix D between the pallets in the picking area and the pallets in the order area: In the matrix D, 1 / (L zisj ) indicates the pallet Z in the picking area i With the pallet S in the order area j The reciprocal of the distance between z1s1 )、1 / (L z2s1 ),...,1 / (L zys1 ) represent the pallets Z1 to Z2 in the picking area y The reciprocal of the distance from the order area pallet S1, and so on; f) Obtain a scoring matrix by multiplying the goods quantity matrix C obtained in step d) by the distance reciprocal matrix D obtained in step e) to obtain the scoring matrix E for all goods with all order numbers on all pallets in the order area: In the matrix E, S ij Indicates that the goods in order number j are sorted to pallet S in the order area i Rating when last used; g). Match the order number and the pallet, find the maximum value of the elements in the matrix E, and let the element with the maximum value in the matrix E be S ij , then the order number j is associated with the pallet S in the order area i Match them together and use order number j with pallet S i The final matching result is stored; at the same time, the matched order number j is deleted from the order number, and the matched pallet S is i Remove from order area tray; h) Determine whether all order numbers have been matched. If so, proceed to step i); if not, proceed to steps d) to g) to match the next order number with the pallet in the order area. i). Goods sorting: After all order numbers are matched, goods are sorted according to the established correspondence between all order numbers and pallets in the order area.

2. The order goods sorting method based on the center-of-gravity greedy idea of ​​block division according to claim 1 is characterized in that: In steps d) and e), if the pallet S in the order area j Since it has established a matching relationship with order number i and is deleted from the order area pallet and order number, the order number i in the picking area pallet Z1 to Z is no longer calculated in step d). y The number of goods on the order area is not calculated in step e) for each pallet in the picking area and the pallet S in the order area. j The distance between them.

3. The order goods sorting method based on the center-of-gravity greedy idea of ​​block division according to claim 1 or 2, characterized in that: The cargo sorting described in step i) is specifically achieved by the following steps: i-1) First, the aerial sorting truss robot identifies the pallet of goods to be sorted that is closest to its current stop in the picking area. It then drives the large and small carts to move the grabbing terminal above the nearest pallet. i-2). Then, scan the barcode on the outer packaging of the goods to identify the goods, and use the grabbing terminal to grab and lift the goods after the barcode is identified; i-3) Next, based on the matching relationship between the order number and the pallet in the order area established in steps g) and h), the pallet number in the order area where the currently grabbed item should be placed is determined. The large and small carts are then driven to move the item to the pallet where it should be placed. The grabbing terminal releases the item and places it on the pallet in the corresponding order area, completing a sorting operation. i-4) After a sorting operation is completed, the aerial sorting truss robot searches for the nearest pallet in the picking area based on its current position and performs the next cargo sorting according to steps i-2) to i-3) until all the cargo in the picking area is sorted to the order area and the sorting task is completed.

Images