Intelligent logistics sorting system based on multi-dimensional recognition and self-adaptive control

Abstract

The invention discloses an intelligent logistics sorting system based on multi-dimensional identification and self-adaptive control, and relates to the field of intelligent logistics sorting, and the system comprises an information input module, the information input module establishes an order database, makes an RFID tag, and inputs order information in the RFID tag; the tree-shaped system establishing module establishes a tree-shaped conveying system, and cargo taking positions are arranged on nodes of the tree-shaped conveying system; the cargo branch combination module is used for calculating the matching degree between the orders, obtaining high-correlation orders and establishing cargo combination branches based on the high-correlation orders; the route roll-out module is used for judging whether the cargoes are boxed or not, and if the cargoes are boxed, the cargoes are rolled out of the boxing route; and a cargo supplementing module. By arranging the tree-shaped system building module, the cargo branch combination module and the route roll-out module, the possibility of manual errors is avoided, and the cargo boxing accuracy is greatly improved.

Description

Technical Field

[0001] The present invention relates to the field of intelligent logistics sorting, and more particularly to an intelligent logistics sorting system based on multi-dimensional recognition and adaptive control. Background Art

[0002] The existence of logistics enables production enterprises to deliver products to the market in a timely manner to meet consumers' demand for products.

[0003] Today, with the rapid development of the logistics industry, logistics sorting faces problems such as a large variety of goods, a large quantity, high labor and time costs for sorting. There are three common logistics sorting methods, namely order picking, batch picking, and combined picking. Each has its own advantages and disadvantages, but they all face the problems of low sorting efficiency and high error rate when there are many types of goods. Summary of the Invention

[0004] To solve the above technical problems, an intelligent logistics sorting system based on multi-dimensional recognition and adaptive control is provided, and this technical solution solves the problems raised in the above background art.

[0005] To achieve the above object, the technical solution adopted by the present invention is as follows:

[0006] An intelligent logistics sorting system based on multi-dimensional recognition and adaptive control includes:

[0007] An information entry module, which establishes an order database, manufactures RFID tags, enters order information into the RFID tags, and based on the order database, equips each order with a corresponding box and attaches the corresponding RFID tag;

[0008] A tree-like system establishment module, which establishes a tree-like transfer system, sets cargo pickup positions at the nodes of the tree-like transfer system, reads the RFID tags on the boxes by an RFID reader, and transmits an electrical signal to a robotic arm, and the robotic arm completes the loading action;

[0009] A cargo branch combination module, which calculates the matching degree between orders, obtains highly correlated orders, and based on the highly correlated orders, establishes a cargo combination branch;

[0010] A path selection module, which when a box passes through each node, the RFID reader reads the RFID tag and selects different paths for the box;

[0011] A route transfer module, which if the cargo is successfully packed, enters relevant information into the corresponding RFID tag, and if the system detects that the box has completed the packing of all cargo, commands the robotic arm to complete sealing the box and transfers the box to the packed route;

[0012] A goods replenishment module, which is provided with a vision sensor to monitor the remaining goods situation at the goods picking position in real time. If the quantity of goods is lower than the preset value, it will direct the relevant staff to replenish the goods;

[0013] The goods branch combination module calculates the matching degree between orders to obtain highly correlated orders. Based on the highly correlated orders, the specific steps for establishing the goods combination branches are as follows:

[0014] Based on the order database, using big data analysis to find highly correlated orders whose matching degree with each other is higher than the preset value. Based on the highly correlated orders, set the corresponding goods combination branches. The goods combination branches refer to that the types of goods at all goods picking positions on the branches of the tree-shaped conveyor system and the matching degree with the highly correlated orders are higher than the preset value. Extract sample orders from the order database, calculate the matching degree between every two orders in the sample orders, summarize all the matching degrees, sort them by size to obtain a matching degree sequence, and take the boundary line of the top 20% of the matching degree sequence to obtain the preset value.

[0015] Preferably, the order information recorded in the RFID tag includes:

[0016] Obtain the identification UID of the RFID tag, and establish a connection between the RFID tag and the order information in the computer;

[0017] The RFID reader reads the identification UID of the RFID tag, and the computer retrieves the order information corresponding to the identification UID;

[0018] The order information includes: the types and quantities of the required goods, the warehousing time, the estimated size of the box, and the logistics transportation method.

[0019] Preferably, for establishing the tree-shaped conveyor system, setting the goods picking positions at the nodes of the tree-shaped conveyor system includes:

[0020] Disperse the goods picking positions, and use the tree-shaped conveyor belt to realize the flow of the boxes between each goods picking position. The tree-shaped conveyor system is composed of the conveyor belts participating in the logistics sorting and transmission. The intersection points of different conveyor belts are the nodes, and the goods picking positions are set at the nodes. The same type of goods is stacked at the goods picking positions;

[0021] Record the number of nodes of the node in the tree-shaped conveyor system from the conveyor entrance as the depth of the node. The depths of the child nodes of the nodes with the same depth are the same. The conveyor entrance is the starting point of the tree-shaped conveyor system;

[0022] Based on the order database, summarize and count the number of times each type of goods appears in all orders, and record it as the appearance times;

[0023] Sort the occurrence times of goods. At the nodes at the same depth, the order of the occurrence times of the goods at at least one goods picking position is the same as the current depth.

[0024] Preferably, based on the order database, high-correlation orders with a matching degree higher than a preset value are analyzed using big data. Based on the high-correlation orders, setting the corresponding goods combination branches includes:

[0025] The number of types of the same goods or the goods placed at the same goods picking position in the high-correlation order group is not less than the preset value;

[0026] An order in which the number of types of the same goods or the goods placed at the same goods picking position in the high-correlation order group reaches the maximum value in the high-correlation order group is called an ultra-high-correlation order group;

[0027] The goods in the corresponding goods combination branch in the high-correlation order group contain the same goods and the goods placed at the same goods picking position in the ultra-high-correlation order group;

[0028] Calculate the matching degree between two orders using the matching degree formula. The matching degree formula is

[0029]

[0030] In the formula, M is the matching degree, x0 is the number of types of the same goods or the goods placed at the same goods picking position between two orders, and x i is the number of types of goods placed at the goods picking positions with a distance difference of i depths between two orders. If the distance of the goods picking position exceeds 3 depths, it does not participate in the counting;

[0031] The number of types of goods in the goods combination branch is the number of types of goods of the order with the largest number of types of goods in the high-correlation order group.

[0032] Preferably, when the box passes through each node, the RFID reader reads the RFID tag. Selecting different paths for the box includes:

[0033] The RFID reader reads the RFID tag on the box to obtain the goods required by the box;

[0034] Input the information of the tree-shaped transfer system into the computer, traverse the nodes of the tree-shaped transfer system. If within n nodes traversed, the goods required by the order are not found, return to the node and search for the route again; if after returning to the node, all routes within n nodes are tried and the goods required by the order are not found, then after returning to the node, traverse all routes within n + 1 nodes and determine whether there are the goods required by the order at the goods picking position of the node, where n starts counting from 1;

[0035] If it is determined that there is the goods required by an order among n nodes and there are different alternative routes, then start traversing from the node with the goods in the order, traverse the routes within m nodes. The node before the two traversals is recorded as the original node, and the node with the goods in the order after the two traversals is recorded as the target point. Search for the path with the shortest distance from the original node to the target point, obtain the path graph, and instruct the path converter on the node to change the path for the box, where m starts counting from 1.

[0036] Preferably, if the goods are successfully packed in the box, the relevant information entered in the corresponding RFID tag includes:

[0037] Based on the number of types of goods in the order, enter the number of types of goods in the order and the base number 0 in the RFID tag;

[0038] Each time the packing is successful, increment the base number 0 in the RFID tag by one;

[0039] If the base number 0 is equal to the number of types of goods in the order, command the robotic arm to complete sealing the box.

[0040] Preferably, if the system detects that the box has completed packing all the goods, command the robotic arm to complete sealing the box and transfer the box to the route for completed packing, including:

[0041] Analyze the size and weight of the goods, obtain the type of the last goods packed, denoted as the last goods, and open a route for completed packing beside the goods picking position where the last goods are placed;

[0042] If there is no directly connected packing route at the node where the box for completed packing is located, plan the shortest path for the box for completed packing to reach the route for completed packing;

[0043] Based on the inbound and outbound order of the order, adjust the order of the boxes on the conveyor belt;

[0044] The reader reads the RFID tag carried by the box and encodes and sorts the boxes based on the inbound and outbound order of the order;

[0045] Select the boxes whose difference in the encoded serial number from the previous box's encoded serial number is higher than the first preset value, and use the robotic arm to place them on the idle platform;

[0046] If it is observed that there is a box on the conveyor belt whose difference in the encoded serial number from the box on the idle platform is lower than the second preset value, put the box on the idle platform back on the conveyor belt.

[0047] Preferably, set a vision sensor to monitor the remaining situation of the goods at the goods picking position in real time. If the quantity of the goods is lower than the preset value, instruct the relevant staff to replenish the goods, including:

[0048] Adjust the position of the vision sensor to ensure that the vision sensor can observe the walls and floors at the cargo picking positions.

[0049] Paint the walls and floors at the cargo picking positions with colors that can be recognized by the vision sensor and are different from the colors of the goods.

[0050] Based on the area of the exposed colors on the walls and floors at the cargo picking positions, determine the quantity of the remaining goods at the cargo picking positions.

[0051] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0052] The hierarchical structure of the tree-shaped conveyor system can flexibly set multiple branches with different combinations of goods compared to a single straight conveyor system. At the intersections of different goods branches, the same goods required by different goods branches can be set, saving space. In addition, by setting goods combination branches with a high matching degree with high-correlation orders, it is convenient for the box to obtain all goods along the shortest path, improving the cargo sorting rate. Finally, the RFID tag of the box is read by the reader and the electrical signal is transmitted to the robotic arm, avoiding the possibility of human error and greatly improving the accuracy of cargo packing. Description of the Drawings

[0053] Figure 1 It is a schematic flow diagram of the intelligent logistics sorting system based on multi-dimensional recognition and adaptive control of the present invention;

[0054] Figure 2 It is a schematic flow diagram of the process of entering order information into the RFID tag of the present invention;

[0055] Figure 3 It is a schematic flow diagram of the process of establishing a tree-shaped conveyor system and setting cargo picking positions at the nodes of the tree-shaped conveyor system of the present invention;

[0056] Figure 4 It is a schematic flow diagram of the process of using big data to analyze high-correlation orders with a matching degree higher than a preset value based on an order database and setting corresponding goods combination branches based on the high-correlation orders of the present invention;

[0057] Figure 5 It is a schematic flow diagram of the process that when the box passes through each node, the RFID reader reads the RFID tag and selects different paths for the box of the present invention;

[0058] Figure 6 It is a schematic flow diagram of the process of entering relevant information into the corresponding RFID tag if the cargo packing is successful of the present invention;

[0059] Figure 7 It is a schematic flow diagram of the process that if the system detects that the box has completed the packing of all goods, it commands the robotic arm to complete the sealing of the box and transfer the box to the route for completed packing of the present invention;

[0060] Figure 8 It is a schematic flow chart of the system for setting up a vision sensor in the present invention to monitor the remaining situation of goods at the real-time position where goods are taken. If the quantity of goods is lower than the preset value, it will command the relevant staff to replenish the goods. Detailed implementation manners

[0061] The following description is used to disclose the present invention so that those skilled in the art can implement the present invention. The preferred embodiments in the following description are only examples, and those skilled in the art can think of other obvious variations.

[0062] Referring to Figure 1 as shown, an intelligent logistics sorting system based on multi-dimensional recognition and adaptive control includes:

[0063] An information input module, which establishes an order database, manufactures RFID tags, and enters order information into the RFID tags. Based on the order database, a corresponding box is allocated for each order and the corresponding RFID tag is pasted on it;

[0064] A tree-like system establishment module, which establishes a tree-like transmission system, sets goods taking positions at the nodes of the tree-like transmission system, reads the RFID tags on the boxes by an RFID reader, and transmits the electrical signals to a robotic arm, and the robotic arm completes the loading action;

[0065] A goods branch combination module, which calculates the matching degree between orders, obtains highly correlated orders, and based on the highly correlated orders, establishes a goods combination branch;

[0066] A path selection module, when the box passes through each node, the RFID reader reads the RFID tag and selects different paths for the box;

[0067] A route transfer-out module, if the goods are successfully packed in the box, it enters relevant information into the corresponding RFID tag. If the system detects that all the goods in the box have been packed, it commands the robotic arm to complete sealing the box and transfers the box to the route for completed packing;

[0068] A goods replenishment module, which sets a vision sensor to monitor the remaining situation of goods at the goods taking position in real time. If the quantity of goods is lower than the preset value, it commands the relevant staff to replenish the goods;

[0069] The goods branch combination module calculates the matching degree between orders, obtains highly correlated orders, and based on the highly correlated orders, establishing a goods combination branch specifically includes the following steps:

[0070] Based on the order database, high - correlation orders with a matching degree higher than a preset value are analyzed using big data. Based on the high - correlation orders, corresponding cargo combination branches are set. The cargo combination branches refer to the types of goods at all cargo - taking positions on the branches of the tree - shaped conveyor system, and the matching degree between the types of goods and the high - correlation orders is higher than the preset value. Sample orders are extracted from the order database, the matching degree between every two orders in the sample orders is calculated, all the matching degrees are summarized and sorted by size to obtain a matching - degree sequence, and the boundary line of the top 20% of the matching - degree sequence is taken to obtain the preset value.

[0071] It can be explained that using RFID tag identification can improve the accuracy of identification, and RFID tags can read multiple tags simultaneously, improving the reading efficiency, thereby improving the efficiency of the entire sorting system. The robotic arm replaces manual labor to receive electrical signals for the boxing operation, avoiding the situation of increased error rate after long - term manual work. The entire system only requires manual placement of boxes with RFID tags attached and appropriate replenishment, and the remaining part realizes full automation, reducing labor costs.

[0072] The application of RFID tags and vision sensors enables the entire system to achieve adaptive control, without the need for manual intervention. Automated boxing and reminder for replenishment are both completed by machines, realizing the intelligence of the entire system.

[0073] Refer to Figure 2 As shown, the order information entered in the RFID tag includes:

[0074] Obtain the identification UID of the RFID tag, and establish a connection between the RFID tag and the order information in the computer.

[0075] The RFID reader reads the identification UID of the RFID tag, and the computer retrieves the order information corresponding to the identification UID.

[0076] The order information includes: the types and quantities of required goods, the warehousing time, the estimated size of the box, and the logistics transportation method.

[0077] It can be explained that summarizing all order data and entering it into the RFID tag facilitates subsequent calling of order data. Utilizing the advantage of convenient reading of RFID tags, it is possible to quickly obtain order information and accurately perform boxing.

[0078] Using the unique identification UID of each RFID tag, a connection is established between the order information and the identification UID, enabling the acquisition of order information by obtaining the RFID tag, improving the convenience of information acquisition.

[0079] Refer to Figure 3 As shown, a tree - shaped conveyor system is established, and cargo - taking positions are set at the nodes of the tree - shaped conveyor system, including:

[0080] The goods picking positions are dispersedly arranged, and the boxes are circulated among the various goods picking positions through a tree-shaped conveyor belt system. The tree-shaped conveyor belt system is composed of conveyor belts participating in logistics sorting and transmission. The intersection points of different conveyor belts are nodes, and the goods picking positions are set at the nodes. The same type of goods is stacked at the goods picking positions.

[0081] The number of nodes from the node of the tree-shaped conveyor belt system to the conveyor entrance is recorded as the depth of the node. The depths of the child nodes of nodes with the same depth are the same. The conveyor entrance is the starting point of the tree-shaped conveyor belt system.

[0082] Based on the order database, the number of times each type of goods appears in all orders is summarized and statistically recorded as the appearance times.

[0083] Sort the appearance times of the goods. At least one of the goods at the goods picking positions at the same depth has the same order of appearance times as the current depth.

[0084] It can be explained that the hierarchical structure of the tree-shaped conveyor belt system can flexibly set multiple branches with different combinations of goods compared with a single straight conveyor belt system. The same goods required by different goods branches can be set at the intersection points of different goods branches, saving space. Place the goods with high appearance frequencies near the conveyor entrance. Since the boxes near the conveyor entrance have high appearance frequencies, most boxes can be packed near the conveyor entrance, improving the packing efficiency.

[0085] Refer to Figure 4 As shown, based on the order database, high-correlation orders with a matching degree higher than a preset value are analyzed using big data. Based on the high-correlation orders, the corresponding goods combination branches are set, including:

[0086] The number of types of the same goods or the goods placed at the same goods picking position in the high-correlation order group is not less than the preset value;

[0087] An order in which the number of types of the same goods or the goods placed at the same goods picking position in the high-correlation order group reaches the maximum value in the high-correlation order group is called an ultra-high-correlation order group;

[0088] The goods in the corresponding goods combination branch in the high-correlation order group contain the same goods and the goods placed at the same goods picking position in the ultra-high-correlation order group;

[0089] The matching degree between two orders is calculated using the matching degree formula. The matching degree formula is

[0090]

[0091] In the formula, M is the matching degree, x0 is the number of types of the same goods or the goods placed at the same goods picking position between the two orders, x iThe number of types of goods at the goods picking location with a distance difference of i depths between two orders. If the distance of the goods picking location exceeds 3 depths, it is not included in the count;

[0092] The number of types of goods in the goods combination branch is the number of types of goods in the order with the largest number of types of goods in the highly correlated order group.

[0093] It can be explained that by calculating the matching degree between orders, a highly correlated order group is obtained, and then a goods combination branch is established based on the highly correlated order group. The goods combinations on the goods combination branch conform to most orders, enabling most orders to be basically packed through one goods combination branch, thus improving the packing efficiency;

[0094] By calculating the matching degree to obtain highly correlated orders, the highly correlated orders achieve setting the goods combination branch with the order group as the target, improving the universality of the goods combination branch.

[0095] Refer to Figure 5 As shown, when the box passes through each node, the RFID reader reads the RFID tag, and the different paths selected for the box include:

[0096] The RFID reader reads the RFID tag on the box to obtain the goods required by the box;

[0097] The information of the tree-shaped transfer system is input into the computer, and the nodes of the tree-shaped transfer system are traversed. If the goods required by the order are not found within n nodes during the traversal, return to the node and search for a route again; if after returning to the node, all routes within n nodes have been tried and the goods required by the order have not been found, then after returning to the node, traverse all routes within n + 1 nodes and determine whether there are the goods required by the order at the goods picking location of the node, where n starts counting from 1;

[0098] If it is determined that there are goods required by the order within n nodes and there are different alternative routes, then start traversing from the node with the goods in the order, traverse the routes within m nodes. The node before the two traversals is recorded as the original node, and the node with the goods in the order after the two traversals is recorded as the target point. Search for the path with the shortest distance from the original node to the target point, obtain the path diagram, and instruct the path converter on the node to change the path for the box, where m starts counting from 1.

[0099] It can be explained that by using the powerful computing power of the computer and adopting a traversal algorithm, the shortest path for the box to obtain goods is obtained. First, determine the shortest path with goods within several nodes, and then screen out the optimal path to reach the next goods location from the shortest paths. The algorithm of "selecting the best from the short" optimizes the convenience of the path.

[0100] Refer to Figure 6 As shown, if the goods are successfully packed, the relevant information is entered into the corresponding RFID tag, including:

[0101] Based on the number of types of goods in the order, enter the number of types of goods in the order and the base number 0 into the RFID tag;

[0102] Each time the boxing is successful, increment the base number 0 in the RFID tag by one;

[0103] If the base number 0 is equal to the number of types of goods in the order, command the robotic arm to complete sealing the box.

[0104] It can be explained that each time the goods are boxed, relevant information is entered into the RFID tag to monitor the goods boxing situation in real time, facilitating the timely transfer to the boxing completed route when the box is being boxed, without occupying the boxing route, which is beneficial to improving the boxing efficiency. By the way of adding numbers rather than comparing that each item of goods is boxed, the computer work is reduced.

[0105] Refer to Figure 7 As shown, if the system detects that the box has completed the boxing of all goods, command the robotic arm to complete sealing the box and transfer the box to the boxing completed route, including:

[0106] Analyze the size and weight of the goods, obtain the type of the last goods boxed, denoted as the last goods, and open a boxing completed route beside the goods picking position where the last goods are placed;

[0107] If there is no directly connected boxing route at the node where the box that has completed boxing is located, plan the shortest path for the box that has completed boxing to reach the boxing completed route;

[0108] Based on the inbound and outbound order of the order, adjust the order of the boxes on the conveyor belt;

[0109] The reader reads the RFID tag carried by the box and sorts the boxes based on the inbound and outbound order of the order;

[0110] Screen out the boxes whose difference in coding serial number from the previous box is higher than the first preset value, and use the robotic arm to place them on the idle platform;

[0111] If it is observed that there is a box on the conveyor belt whose difference in coding serial number from the box on the idle platform is lower than the second preset value, put the box on the idle platform back onto the conveyor belt.

[0112] It can be explained that sorting the boxes based on the inbound and outbound order of the order, taking out the boxes with serious misalignment and placing them on the idle platform. By this method, the order of the boxes is adjusted, facilitating the boxes to be arranged in the correct inbound and outbound order, and facilitating subsequent personnel to pick up and organize the boxes;

[0113] By opening a route for completed packing beside the cargo picking position where the last cargo is placed, it is ensured that most boxes can directly enter the completed packing route after packing is completed, reducing the stay of the packed boxes on the packing route and optimizing the occupancy of boxes on the path.

[0114] Refer to Figure 8 As shown, a vision sensor is set to monitor the remaining situation of the cargo at the cargo picking position in real time. If the quantity of the cargo is lower than the preset value, relevant staff will be directed to replenish the cargo, including:

[0115] Adjust the position of the vision sensor to ensure that the vision sensor can observe the walls and floors of the cargo picking position;

[0116] Paint the walls and floors of the cargo picking position with colors that can be recognized by the vision sensor and are different from the colors of the cargo;

[0117] Based on the area of the exposed color on the walls and floors of the observed cargo picking position, judge the quantity of the remaining cargo at the cargo picking position.

[0118] It can be explained that by painting colors on the floor and walls, the computer analyzes the painted color area of the pictures obtained by the vision sensor to obtain the remaining situation of the cargo in the cargo picking area. Compared with directly analyzing the quantity of the cargo, it can reduce the computing power of the computer.

[0119] Furthermore, this solution also proposes a storage medium on which a computer-readable program is stored. When the computer-readable program is called, the above-mentioned intelligent logistics sorting system based on multi-dimensional recognition and adaptive control runs.

[0120] It can be understood that the storage medium can be a magnetic medium, such as a floppy disk, a hard disk, or a magnetic tape; an optical medium, such as a DVD; or a semiconductor medium, such as a solid-state drive (SSD).

[0121] In summary, the advantages of the present invention are as follows: The hierarchical structure of the tree-shaped conveyor system can flexibly set multiple branches with different cargo combinations compared with a single straight conveyor system. The same cargo required by different cargo branches can be set at the intersection of different cargo branches, saving space. In addition, cargo combination branches with a high matching degree with high-correlation orders are set, which is convenient for the boxes to obtain all the cargo with the shortest path and improves the cargo sorting rate. Finally, the RFID tag of the box is read by the reader and the electrical signal is transmitted to the robotic arm, avoiding the possibility of human error and greatly improving the accuracy of cargo packing.

[0122] The foregoing has shown and described the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited by the above embodiments, and what is described in the above embodiments and the specification is only the principle of the present invention. Without departing from the spirit and scope of the present invention, various changes and improvements will occur to the present invention, and these changes and improvements fall within the scope of the present invention claimed. The scope of protection required by the present invention is defined by the appended claims and their equivalents.

Claims

1. An intelligent logistics sorting system based on multi-dimensional recognition and adaptive control, characterized in that Including: An information entry module, which establishes an order database, manufactures RFID tags, enters order information into the RFID tags, and based on the order database, assigns a corresponding box to each order and attaches the corresponding RFID tag; A tree - like system establishment module, which establishes a tree - like transportation system, sets cargo picking positions at the nodes of the tree - like transportation system, reads the RFID tag on the box by an RFID reader, and transmits an electrical signal to a robotic arm, and the robotic arm completes the loading operation; A cargo branch combination module, which calculates the matching degree between orders, obtains highly correlated orders, and based on the highly correlated orders, establishes cargo combination branches; A path selection module, which when the box passes through each node, the RFID reader reads the RFID tag and selects different paths for the box; A route transfer module, which if the cargo loading is successful, enters relevant information into the corresponding RFID tag, and if the system detects that the box has completed the loading of all cargo, commands the robotic arm to complete sealing the box and transfers the box to the route for completed loading; A cargo replenishment module, which sets a vision sensor to monitor the remaining situation of the cargo at the cargo picking position in real - time, and if the quantity of the cargo is lower than a preset value, commands the relevant staff to replenish the cargo; The specific steps for the cargo branch combination module to calculate the matching degree between orders, obtain highly correlated orders, and establish cargo combination branches based on the highly correlated orders are as follows: Based on the order database, using big data analysis to find highly correlated orders whose matching degree with each other is higher than a preset value, and based on the highly correlated orders, set corresponding cargo combination branches. The cargo combination branches refer to that the types of cargo at all cargo picking positions on the branches of the tree - like transportation system and the matching degree of the highly correlated orders are higher than the preset value. Extract sample orders from the order database, calculate the matching degree between every two orders in the sample orders, summarize all the matching degrees, sort them by size to obtain a matching degree sequence, and take the boundary line of the top twenty percent of the matching degree sequence to obtain the preset value.

2. The intelligent logistics sorting system based on multi-dimensional recognition and adaptive control according to claim 1, wherein, The order information entered into the RFID tag includes: Obtain the identification UID of the RFID tag and establish a connection between the RFID tag and the order information in the computer; The RFID reader reads the identification UID of the RFID tag, and the computer retrieves the order information corresponding to the identification UID; The order information includes: the type and quantity of the required cargo, the warehousing time, the estimated size of the box, and the logistics transportation method.

3. An intelligent logistics sorting system based on multi-dimensional recognition and adaptive control according to claim 2, characterized in that, The steps for establishing the tree - like transportation system and setting cargo picking positions at the nodes of the tree - like transportation system include: Disperse the setting of cargo picking positions, and use a tree - like conveyor belt to realize the flow of the box between each cargo picking position. The tree - like transportation system is composed of conveyor belts participating in logistics sorting and transmission, and the intersection points of different conveyor belts are nodes. The cargo picking positions are set at the nodes, and the same type of cargo is stacked at the cargo picking positions. In the tree-shaped transfer system, the number of nodes from a node to the transfer entrance is recorded as the depth of the node. The children nodes of nodes with the same depth have the same depth, and the transfer entrance is the starting point of the tree-shaped transfer system. Based on the order database, summarize and count the number of times each type of goods appears in all orders, which is recorded as the appearance times. Sort the appearance times of the goods. At least one goods picking position at nodes with the same depth has the goods appearance times order the same as the current depth.

4. An intelligent logistics sorting system based on multi-dimensional recognition and adaptive control according to claim 3, characterized in that, Based on the order database, use big data to analyze highly correlated orders with a matching degree higher than a preset value. Based on the highly correlated orders, set the corresponding goods combination branches including: The number of types of the same goods or goods placed at the same goods picking position in the highly correlated order group is not less than the preset value. An order in the highly correlated order group where the number of types of the same goods or goods placed at the same goods picking position reaches the maximum value in the highly correlated order group is called an ultra-highly correlated order group. The goods in the corresponding goods combination branch in the highly correlated order group contain the same goods in the ultra-highly correlated order group and the goods placed at the same goods picking position. Calculate the matching degree between two orders using the matching degree formula, and the matching degree formula is Wherein, M is the matching degree, x0 is the number of types of goods that are the same between two orders or the goods placed at the same goods picking position, and xi i is the number of types of goods placed at the goods picking positions with a depth difference of i between two orders. If the distance between the goods picking positions exceeds 3 depths, it is not included in the count; The number of types of goods in the goods combination branch takes the number of types of goods in the order with the largest number of types of goods in the highly correlated order group.

5. An intelligent logistics sorting system based on multi-dimensional recognition and adaptive control according to claim 4, characterized in that When the box passes through each node, the RFID reader reads the RFID tag and selects different paths for the box, including: The RFID reader reads the RFID tag on the box to obtain the goods required by the box. Input the information of the tree-shaped transfer system into the computer, traverse the nodes of the tree-shaped transfer system. If within n nodes traversed, the goods required by the order are not found, return to the node and search for the route again; if after returning to the node, all routes within n nodes are tried and the goods required by the order are not found, then after returning to the node, traverse all routes within n + 1 nodes, and judge whether there are the goods required by the order at the goods picking position of the node, where n starts counting from 1. If it is judged that there are the goods required by the order within n nodes and there are different alternative routes, start traversing from the node with the goods in the order, traverse the routes within m nodes. The node before the two traversals is recorded as the original node, and the node with the goods in the order after the two traversals is recorded as the target point. Search for the path with the shortest distance from the original node to the target point, obtain the path diagram, and instruct the path converter on the node to change the path for the box, where m starts counting from 1.

6. An intelligent logistics sorting system based on multi-dimensional recognition and adaptive control according to claim 5, characterized in that If the goods are successfully packed in the box, the relevant information is recorded in the corresponding RFID tag, including: Based on the number of types of goods in the order, record the number of types of goods in the order and the base number 0 in the RFID tag. Each time the packing is successful, add 1 to the base number 0 in the RFID tag. If the base number 0 is equal to the number of types of goods in the order, command the robotic arm to complete sealing the box.

7. An intelligent logistics sorting system based on multi-dimensional recognition and adaptive control according to claim 6, characterized in that, If the system detects that the box has completed packing all the goods, command the robotic arm to complete sealing the box and transfer the box to the packed route, including: Analyze the size and weight of the goods, obtain the type of the last goods packed, which is recorded as the last goods, and open a packed route beside the goods picking position where the last goods is placed. If there is no directly connected packing route for the node where the box that has been packed is located, plan the shortest path for the box that has been packed to reach the packing completion route; Based on the inbound and outbound order of the order, adjust the order of the boxes on the conveyor belt; The reader reads the RFID tag carried by the box and sorts the boxes based on the inbound and outbound order of the order; Screen out the boxes whose difference in coding serial number from the previous box's coding serial number is higher than the first preset value, and use the robotic arm to place them on the idle platform; If it is observed that the difference in coding serial number between the box on the conveyor belt and the box on the idle platform is lower than the second preset value, then put the box on the idle platform back onto the conveyor belt.

8. An intelligent logistics sorting system based on multi-dimensional recognition and adaptive control according to claim 7, characterized in that, The visual sensor is set to monitor the remaining situation of the goods at the goods picking position in real time. If the quantity of goods is lower than the preset value, direct the relevant staff to replenish the goods, including: Adjust the position of the visual sensor to ensure that the visual sensor can observe the wall and floor at the goods picking position; Paint the wall and floor at the goods picking position with a color that can be recognized by the visual sensor and is different from the color of the goods; Based on the area of the exposed color on the wall and floor at the goods picking position, judge the quantity of the remaining goods at the goods picking position.

Images