An automatic transportation planning method for intelligent three-dimensional storage of finished grains

Abstract

The present invention relates to the technical field of warehouse management, and in particular to an automatic transportation planning method for intelligent three-dimensional storage of finished grains, the method comprising: obtaining the position of a transport vehicle and the category of goods loaded on the transport vehicle in real time, and taking the position as a starting point; confirming the end point for storing the goods based on the category; planning the travel path of the goods according to the starting point and end point of the goods, and completing the transportation management of the goods; that is, the scheme of the present invention can improve the accuracy of path planning in warehouse management.

Description

Technical Field

[0001] The present invention relates to the field of warehouse management technology, and more specifically, to an automatic transportation planning method for intelligent three-dimensional storage of finished grains. Background Art

[0002] Finished grains usually refer to grain products that can be eaten directly after processing, such as rice, flour, edible vegetable oil, etc. Finished grains are the most basic living needs of residents. Finished grain reserves are the most direct and effective means of regulating the grain market. Therefore, finished grain storage management is of great significance to ensuring food security.

[0003] The existing intelligent three-dimensional storage system can not only vertically increase the storage space and make full use of the storage space, but also realize the automatic unloading, stacking, transportation, storage and retrieval, information scanning and recording of packaged finished products through fully automatic conveying equipment, which greatly reduces the waiting time for operations, has high work efficiency, low labor costs, and greatly improves the safety of operations. At the same time, intelligent three-dimensional storage can also ensure a low-temperature storage environment by setting up a low-temperature grain storage unit, using duct temperature equalization technology to reduce the temperature difference of the entire storage warehouse, greatly reducing rice weevils, condensation and cold core phenomena, and realizing safe storage of finished grain.

[0004] Among them, path planning during cargo transportation is particularly important in warehouse management, and there are many methods for path planning, such as graph theory and shortest path algorithm; in related technologies, such as the patent application document with application publication number CN118485378A and name Intelligent Grain Storage System and Method, which discloses the use of graph theory and shortest path algorithm to calculate the best path for an automated guided vehicle (AGV), taking into account the priority of the cargo and the hardness of the handling equipment, and optimizing the path; but the disadvantage of this algorithm is that when there are negative weight edges in the graph, it may not be able to find the shortest path or may not be able to terminate at all.

[0005] Another example: The (Rapidly-exploring Random Trees Star) algorithm is an algorithm for efficient path planning, especially for automatic guided vehicles (AGVs). It includes the steps of initialization, random sampling, tree expansion, reselection of parent nodes, node reconnection, path optimization, and determination of termination conditions. In the tree expansion step, the conventional The search step size in the algorithm is fixed, which may not adapt to complex environments and result in the inability to obtain an accurate driving path.

[0006] Therefore, how to obtain the accurate driving route is particularly important for the storage management of finished grains. Summary of the invention

[0007] The purpose of the present invention is to propose an automatic transportation planning method for intelligent three-dimensional storage of finished grains, so as to solve the problem of inaccuracy in path planning in the prior art; to this end, the present invention provides a solution in the following aspect.

[0008] The present invention provides an automatic transportation planning method for intelligent three-dimensional storage of finished grains, comprising:

[0009] Get the current location and category of the goods in real time, using the current location as the starting point;

[0010] Identify the destination of the goods based on their type;

[0011] According to the starting point and the end point, the improved The algorithm plans the route of the goods to complete the transportation management of the goods;

[0012] Among them, the improved The algorithm includes an adaptive search step size in the tree expansion step The calculation is as follows: ; is the maximum feasible distance of the current node along the search direction, is the angle between the current search vector and the target vector; is the density of obstacles at the current node, is a cosine function; the maximum feasible distance is the distance traveled by the transport vehicle when it first collides with an obstacle when traveling along the search direction at the current node; the target vector is a vector pointing from the starting point to the end point, and the current search vector is a unit vector constructed along the search direction with the current node; the search direction is an arbitrary direction with the current node as the starting point.

[0013] In the above scheme, the adaptive search step length of the current node along the search direction is calculated by the maximum feasible distance, the density of obstacles, and the angle between the search vector and the target vector, which can improve the accuracy of the planning of the driving path.

[0014] Optionally, the obstacle density for:

[0015] ;in, and Respectively represent the length and width of the transport vehicle; represents the total number of obstacles within the circle, Indicates the circle The total area of ​​obstacles; the circle is centered on the centroid of the transport vehicle, and the length of the diagonal of the transport vehicle is times the radius, where >5.

[0016] In the above scheme, the situation of obstacles around the transport vehicle can be obtained.

[0017] Optionally, the tree expansion step is specifically:

[0018] Find the nearest neighbor node Xnear in the constructed random tree, and determine the path between any sampling point Xrand of the nearest neighbor node Xnear;

[0019] Check whether the path between the nearest neighbor node Xnear and any sampling point Xrand collides with an obstacle. If a collision occurs, return the improved The random sampling step in the algorithm; if there is no collision, the nearest neighbor node Xnear is used as the starting point, and the search direction is extended according to the adaptive search step size to generate a new node Xnew, connect the new node Xnew and the nearest neighbor node Xnear, and continue to determine whether the path between the new node Xnew and the nearest neighbor node Xnear collides to expand the random tree.

[0020] Optionally, the improved The algorithm also includes the movement cost in the node reconnection step The calculation is as follows:

[0021] ;

[0022] in, Represents the total number of nodes from the starting point along the path in the random tree to the current node. Representation Node To Node The Euclidean distance of Represented by the node Point to Node The determined vectors and nodes Point to Node The angle between the vectors determined, for The maximum obstacle density among the nodes.

[0023] The above scheme not only ensures the accuracy of the path movement cost calculation by introducing the turning angles and obstacle density between different nodes, but also considers the safety of the path and avoids potential safety risks.

[0024] Optionally, the specific process of the node reconnection step is: after the new node Xnew is added to the random tree, check all neighborhood nodes of the new node Xnew, calculate the movement cost from the new node Xnew to each neighborhood node, determine the neighborhood node with the minimum movement cost, update the parent node of the neighborhood node with the minimum movement cost to the new node Xnew, and reconnect the nodes.

[0025] Optionally, the method further includes the step of determining whether the driving paths of the multiple transport vehicles are congested or collided, specifically:

[0026] Determine the intersection of multiple travel paths;

[0027] Calculate the time it takes for the transport vehicle on the path corresponding to each intersection to arrive at the corresponding intersection;

[0028] Arrange the times in order, calculate the difference between two adjacent times, and when the difference is greater than a threshold, the transport vehicles at the two times will not collide or get stuck;

[0029] When the difference is less than or equal to a threshold, the transport vehicles at the two times collide or become jammed.

[0030] The above scheme provides a basis for subsequent measures by judging congestion or collision at the intersection of the initial paths of multiple transport vehicles.

[0031] Optionally, when a collision or congestion occurs among vehicles, the method further includes: controlling vehicles with lower priorities to slow down or stop according to the priorities of the transport vehicles.

[0032] The above scheme adjusts the traveling speed of congested vehicles to avoid vehicle collision or congestion.

[0033] Optionally, the specific process of confirming the destination of storing goods based on the category of goods is:

[0034] The storage location of the corresponding goods is queried according to the category of the goods, and the storage location is used as the end point of storing the goods; the category of the goods is rice, flour or edible vegetable oil.

[0035] The beneficial effects of the present invention are:

[0036] The solution of the present invention introduces the maximum feasible distance, the density of obstacles, and the angle between the search vector and the target vector to improve the existing The search step size in the algorithm is adjusted dynamically to improve the efficiency of path search and reduce costs. At the same time, by introducing the Euclidean distance between adjacent nodes in the path, the turning angle between nodes, and the density of obstacles at the nodes, the movement cost of the transport vehicle from the starting point to the current node is calculated, which can improve the accuracy and safety of the path cost and thus obtain an accurate driving path. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] By reading the following detailed description with reference to the accompanying drawings, the above and other objects, features and advantages of the exemplary embodiments of the present invention will become readily understood. In the accompanying drawings, several embodiments of the present invention are shown in an exemplary and non-limiting manner, and the same or corresponding reference numerals represent the same or corresponding parts, wherein:

[0038] Figure 1 A flowchart schematically shows a method for automatic transportation planning of finished grain intelligent three-dimensional storage in this embodiment;

[0039] Figure 2 A schematic diagram schematically shows the maximum drivable distance between the transport vehicle and the obstacle in this embodiment;

[0040] Figure 3 The schematic diagram of the driving path of the transport vehicle in this embodiment is schematically shown. DETAILED DESCRIPTION

[0041] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0042] The present invention is aimed at the management of path planning in the storage management of finished grains, which is generally divided into two types. One is that the transport vehicle needs to go to the destination to pick up goods. At this time, the transport vehicle is empty, the location of the transport vehicle is the starting point, and the location of the goods is the destination; the other is that the transport vehicle is loaded with goods and needs to deliver the goods to the destination. At this time, the location of the transport vehicle is the starting point, and the storage location of the goods is the destination.

[0043] Specifically, take the example of loading cargo on a transport vehicle. Figure 1 As shown, an automatic transportation planning method for finished grain intelligent three-dimensional storage in this embodiment includes the following steps:

[0044] Step S1, obtaining the position of the transport vehicle and the type of goods loaded on the transport vehicle in real time, and taking the position as the starting point.

[0045] The above-mentioned transport vehicle is an automatic guided vehicle (AGV); the types of goods are rice, flour or edible vegetable oil among finished grains.

[0046] In this embodiment, the position of the transport vehicle can be monitored by devices such as a global satellite positioning system and a wireless sensor network.

[0047] Step S2, confirming the destination where the goods need to be stored based on the category. Specifically, according to the category of the goods, query the storage location corresponding to the goods, and use the queried storage location as the destination for storing the goods.

[0048] The storage location corresponding to the above-mentioned queried goods can be determined based on whether the category of the goods corresponds one-to-one with the storage locations of goods of different categories stored.

[0049] Taking flour as an example, the ideal storage temperature is 18 to 24 degrees (to prevent the flour from deteriorating due to high temperature). High temperature and humid environment can easily cause flour to deteriorate, so the storage location needs to be set in a clean, dry and ventilated storage environment.

[0050] Step S3, according to the starting point and the end point, adopt the improved The algorithm plans the travel route of the goods to complete the transportation management of the goods.

[0051] in, The Rapidly-exploring Random Trees Star (RRT) algorithm is a search-based path planning algorithm that finds the optimal solution by randomly sampling points in the environment and gradually building a path. It can handle complex dynamic environments and has good real-time performance. The main feature of the RRT algorithm is that it can quickly find the initial path, and then continuously optimizes as the number of sampling points increases until the end point is found or the set maximum number of cycles is reached.

[0052] In this embodiment, the improved The steps of the algorithm to plan the travel path of goods are as follows:

[0053] 1. Initialization: Initialize the starting point and the end point, take the starting point as the root node, and build a random tree that only contains the starting point. Specifically, the location of the transport vehicle loaded with goods in the warehouse is called the starting point, and the final destination of the goods to be transported is called the end point.

[0054] 2. Repeat the following steps af until the driving path is found or the maximum number of iterations is reached and the expansion stops. The specific steps are:

[0055] Step a, random sampling: establish a sampling space according to the starting point and the end point, and randomly sample any sampling point Xrand in the sampling space.

[0056] Step b, expanding the tree: finding the nearest neighbor node Xnear in the random tree, and determining the path between the nearest neighbor node Xnear and any sampling point Xrand;

[0057] Check whether the path from the nearest neighbor node Xnear to any sampling point Xrand collides with an obstacle. If a collision occurs, return to step a and continue the next iteration. If there is no collision, take the nearest neighbor node Xnear as the starting point, extend it in the search direction according to the adaptive search step size, generate a new node Xnew, connect the new node Xnew with the nearest neighbor node Xnear, and continue to determine whether the path between the new node Xnew and the nearest neighbor node Xnear collides, so as to expand the random tree.

[0058] The nearest neighbor node Xnear is the point where the distance between all nodes in the random tree and any sampling point Xrand is the smallest. The search direction can be any direction starting from the nearest neighbor node Xnear. The specific process of the collision detection is prior art and will not be described in detail here.

[0059] The process of obtaining the above adaptive search step size is:

[0060] First, a target vector is constructed with the starting point and the end point, and the direction of the target vector is from the starting point to the end point; each node corresponding to the adaptive search step to be analyzed is called the current node; a unit vector is constructed with the search direction of the current node and is used as the current search vector, with the direction being along the search direction.

[0061] Secondly, at the current node, along the search direction, the maximum feasible distance of the transport vehicle is calculated; the maximum feasible distance refers to the distance traveled by the transport vehicle when it first collides with an obstacle while traveling along the search direction at the current node, such as Figure 2 shown.

[0062] Then, the obstacle density at the current node is calculated.

[0063] In one embodiment, the above obstacle density for:

[0064] ;in, and Respectively represent the length and width of the transport vehicle; represents the total number of obstacles within the circle, Indicates the circle The total area of ​​obstacles; the circle is centered on the transport vehicle, and the length of the diagonal of the transport vehicle is times the radius, where >5.

[0065] For example, ; Of course, as other implementation methods, It can also be determined according to actual conditions. is the area of ​​the entire circle, Indicates the area of ​​the obstacle within the circle. The larger the value, the denser the obstacles are at the current node. The larger the obstacle area in the circle, the The smaller it is, the denser the obstacles are at the current node. The smaller.

[0066] Finally, the adaptive search step length of the current node is calculated based on the maximum feasible distance and the density of obstacles. , specifically: ; is the maximum feasible distance of the current node along the search direction, is the angle between the current search vector and the target vector; is the density of obstacles at the current node, is the cosine function.

[0067] Among them, when the maximum feasible distance The larger it is, the farther the obstacle in the search direction is. should be larger; when the maximum feasible distance The smaller it is, the closer the obstacle in the search direction is. In order to avoid collision, The smaller the The larger the value, the greater the deviation between the current search direction and the target vector direction. When the current search direction is opposite to the direction of the target vector, the further away from the end point the The smaller the The smaller it is, the smaller the deviation between the current search direction and the target vector is, and the closer it is to the end point. The larger the value, the faster the search speed. The larger the value, the more obstacles there are around. Should be smaller to avoid collision; obstacle density The smaller it is, the more open the surrounding area is and the fewer obstacles there are. The larger the value, the faster the search.

[0068] The reason for setting the adaptive search step size is that when the step size is too small, the exploration efficiency may decrease, the computational cost may increase, and a small step size may generate more nodes, increase the complexity of the tree, and increase the computational cost of subsequent reconnection operations and optimization paths; when the step size is too large, it is easy to miss the target or the optimal path, and the path optimization is difficult, and the path generated by a large step size may be rough, requiring more subsequent optimization steps to smooth the path, which increases the difficulty of path optimization; at the same time, a large step size may also cause the new node to fall directly on or near an obstacle, increasing the complexity of collision detection. Therefore, by setting a suitable adaptive search step size, the needs of different nodes can be met.

[0069] Step c, reselect the parent node (find the optimal connection): find the nearest node Xmin of the new node Xnew in the random tree, take the nearest node Xmin as the parent node of the new node Xnew to update the parent node of the new node Xnew, and calculate the path cost from the nearest node Xmin to the new node Xnew.

[0070] The path cost mentioned above is the Euclidean distance of the path between two nodes; since it is an existing technology, it will not be described here.

[0071] Step d, node reconnection: After the new node Xnew is added to the random tree, check all the neighborhood nodes of the new node Xnew, calculate the movement cost from the new node Xnew to each neighborhood node, determine the neighborhood node with the minimum movement cost, update the parent node of the neighborhood node with the minimum movement cost to the new node Xnew, and reconnect the nodes.

[0072] The nodes in the neighborhood are obtained by searching on a random tree with the new node Xnew as the center and r as the radius, wherein the value range of r is [0.1, 0.5]; other implementation methods can also be determined according to actual conditions.

[0073] The moving cost in this example for:

[0074] ; Represents the total number of nodes from the starting point along the path in the random tree to the current node. Representation Node To Node The Euclidean distance of Represented by the node Point to Node The determined vectors and nodes Point to Node The angle between the vectors determined, for The maximum obstacle density among the nodes.

[0075] in, It is to Normalized to the range of (0, 1), It is the average of the angles (turning angles) between all nodes. The larger the turning angle, the more curved and complex the path is, which indirectly increases the total path length from the starting point to the end point. This is because the straight line between two points is the shortest, and the more curved it is, the longer the total path may be, thereby increasing the movement cost.

[0076] For example, Figure 3 As shown, For Node To Node The Euclidean distance of Represented by the node Point to Node The determined vectors and nodes Point to Node The angle of the determined vector can be used to calculate the movement cost of the path from the starting point to the current node.

[0077] The reason for introducing the density of obstacles is that when the obstacles at a certain node are particularly dense, that is, The larger it is, the greater the density of obstacles when multiple transport vehicles are carrying out transportation tasks at the same time, the smaller the width of the feasible path at that location, and the greater the potential possibility of collision and congestion of multiple transport vehicles, thereby increasing safety hazards and indirectly leading to a decrease in transportation efficiency, making the movement cost of the path greater.

[0078] Each time a new node is generated, the movement cost from the starting point to the new node needs to be calculated. When calculating the movement cost, by introducing two factors, the angle and the density of obstacles, the accuracy and safety of the acquired path can be improved.

[0079] Step e, optimizing the path: optimizing some nodes in the random tree with the end point as the target, and improving the path quality by adjusting the path relationship and the moving cost. This step is not described in detail because it is an existing technology.

[0080] Step f, determine the termination condition; if the added node is close to the end point, a feasible driving path is found and the algorithm ends.

[0081] For example, Figure 3 Among them, X0 is the starting point, that is, the location of the goods and the transport vehicle, B is the end point, that is, the storage location of the goods; X1, X2, X3 and X4 are nodes on the path.

[0082] The present invention only applies to existing The search step length and movement cost in the algorithm are improved, and the other steps remain unchanged. Compared with conventional algorithms, this algorithm has high search efficiency, accurate potential cost calculation, and can obtain accurate driving paths.

[0083] Furthermore, after determining the driving paths of the transport vehicles corresponding to each cargo, the method further includes the step of determining whether the driving paths of the multiple transport vehicles are congested or collided, specifically:

[0084] Determine the intersection of multiple driving paths; for any intersection, calculate the time it takes for a transport vehicle on the path corresponding to the intersection to arrive at the intersection; arrange the times in chronological order, and calculate the difference between two adjacent times. When the difference is greater than a threshold, the transport vehicles corresponding to the two times will not collide or get stuck; when the difference is less than or equal to the threshold, the transport vehicles corresponding to the two times may collide or get stuck.

[0085] Among them, when the vehicles collide or get stuck, it also includes: according to the priority of the transport vehicle, controlling the vehicles with lower priority to slow down or stop.

[0086] The above priority is determined according to the priority of the goods, which can be specifically determined according to the requirements of the storage environment. For example, goods with high requirements for the storage environment have a high priority (such as flour), and goods with slightly lower requirements for the storage environment have a low priority (such as edible vegetable oil). Of course, it can also be considered to set priorities.

[0087] The above time is determined according to the path length and driving speed.

[0088] The threshold value may be 10s; of course, as other implementations, it may also be set according to actual conditions.

[0089] The reason for the above-mentioned judgment of collision or congestion between the driving paths of multiple transport vehicles is that there are usually multiple transport vehicles transporting goods at the same time in the warehouse. Therefore, after obtaining the driving paths of multiple transport vehicles, the possibility of traffic congestion occurring at the intersection of different paths needs to be calculated to improve the safety of transportation.

[0090] The solution of the present invention fully considers the complex environment, dense shelves and many obstacles in the warehouse, and sets an adaptive search step length to adjust the existing Improvements to the algorithm can improve search efficiency and calculate the maximum feasible distance for each search direction to ensure that the generated path is actually passable by vehicles. At the same time, in the calculation of the movement cost in the node reconnection step, the steering angle and the density of obstacles at the node are added, which not only makes the movement cost calculation of the path more accurate, but also considers the safety of the path, avoids potential safety risks, and finally plans a safe and low-cost path for the transportation of finished grain.

[0091] Although this specification has shown and described a number of embodiments of the present invention, it will be apparent to those skilled in the art that such embodiments are provided by way of example only. Those skilled in the art will conceive of many modifications, changes and alternatives without departing from the ideas and spirit of the present invention. It should be understood that in the practice of the present invention, various alternatives to the embodiments of the present invention described herein may be employed.

Claims

1. An automatic transportation planning method for intelligent three-dimensional storage of finished grains, characterized in that: include: Acquire the location of the transport vehicle and the type of cargo loaded on the transport vehicle in real time, and use the location as a starting point; Identify the destination of the goods based on their type; According to the starting point and the end point, the improved The algorithm plans the route of the goods to complete the transportation management of the goods; Among them, the improved The algorithm includes an adaptive search step size in the tree expansion step The calculation of and movement cost in the node reconnection step The calculation is as follows: ; is the maximum feasible distance of the current node along the search direction, is the angle between the current search vector and the target vector; is the density of obstacles at the current node, and its calculation formula is: , and Respectively represent the length and width of the transport vehicle, represents the total number of obstacles within the circle, Indicates the circle The total area of ​​obstacles, the circle is centered on the centroid of the transport vehicle, and the length of the diagonal of the transport vehicle is times the radius, where >5; is a cosine function; the maximum feasible distance is the distance traveled by the transport vehicle when it first collides with an obstacle when traveling along the search direction at the current node; the target vector is a vector pointing from the starting point to the end point, and the current search vector is a unit vector constructed along the search direction with the current node; the search direction is an arbitrary direction with the current node as the starting point; ; Represents the total number of nodes from the starting point along the path in the random tree to the current node. Representation Node To Node The Euclidean distance of Represented by the node Point to Node The determined vectors and nodes Point to Node The angle between the vectors determined, for The maximum obstacle density among the nodes; The specific steps of expanding the tree are as follows: find the nearest neighbor node Xnear in the constructed random tree, determine the path between the nearest neighbor node Xnear and any sampling point Xrand; detect whether the path between the nearest neighbor node Xnear and any sampling point Xrand collides with an obstacle, and if a collision occurs, return the improved The random sampling step in the algorithm; if there is no collision, the nearest neighbor node Xnear is used as the starting point, and the search direction is extended according to the adaptive search step size to generate a new node Xnew, connect the new node Xnew and the nearest neighbor node Xnear, and continue to determine whether the path between the new node Xnew and the nearest neighbor node Xnear collides to expand the random tree.

2. According to claim 1, the automatic transportation planning method for finished grain intelligent three-dimensional storage is characterized in that: The specific process of the node reconnection step is as follows: after the new node Xnew is added to the random tree, all the neighborhood nodes of the new node Xnew are checked, the movement cost from the new node Xnew to each neighborhood node is calculated, the neighborhood node with the minimum movement cost is determined, the parent node of the neighborhood node with the minimum movement cost is updated to the new node Xnew, and the nodes are reconnected.

3. The automatic transportation planning method for finished grain intelligent three-dimensional storage according to claim 1 is characterized in that: The method also includes the step of determining whether the driving paths of the multiple transport vehicles are congested or collided, specifically: Determine the intersection of multiple travel paths; Calculate the time it takes for the transport vehicle on the path corresponding to each intersection to arrive at the corresponding intersection; Arrange the times in chronological order and calculate the difference between two adjacent times. When the difference is greater than a threshold, the transport vehicles corresponding to the two times will not collide or get stuck; when the difference is less than or equal to the threshold, the transport vehicles corresponding to the two times will collide or get stuck.

4. The automatic transportation planning method for finished grain intelligent three-dimensional storage according to claim 3 is characterized in that: When vehicles collide or get stuck in traffic, the method also includes: controlling vehicles with lower priorities to slow down or stop according to the priorities of the transport vehicles.

5. The automatic transportation planning method for finished grain intelligent three-dimensional storage according to claim 4 is characterized in that: The specific process of confirming the destination of storing goods based on the category of goods is as follows: The storage location of the corresponding goods is queried according to the category of the goods, and the storage location is used as the end point of storing the goods; the category of the goods is rice, flour or edible vegetable oil.

Images