Unmanned forklift and processing unit thereof

Abstract

The invention relates to an unmanned forklift and a processing unit thereof, and the processing unit is configured as follows: a minimum cube capable of completely containing goods is used as a bounding box, and a virtual channel capable of representing the motion trail of the unmanned forklift is established in a mode that the bounding box moves along a preset carrying route; whether the unmanned forklift collides with the obstacle or not under the carrying route is judged through pre-collision of the virtual channel and the bounding box, and therefore the carrying route suitable for the cargoes and the unmanned forklift to move is planned. According to the method, the bounding box to be detected corresponding to the unmanned forklift is screened out in the area with the bounding box for collision inspection, and the efficiency of data calculation is improved. According to the method, the early-stage calculation amount of preset path collision prediction is reduced, and the safe carrying route of the unmanned forklift can be planned.

Description

[0001] The original basis of this divisional application is a patent application with application number (202410059352.8), application date 2024.01.15, and invention name “A cargo handling route planning system and method”. Technical Field

[0002] The present invention relates to the technical field of intelligent forklifts, and in particular to an unmanned forklift and a processing unit thereof. Background Art

[0003] A forklift is an industrial handling vehicle, which refers to a handling vehicle that loads, unloads, stacks, and transports goods over short distances. Forklift technology is changing with each passing day, especially the intelligence of forklifts. When an unmanned forklift is driving or forking goods in a workshop or high-risk environment, the placement of some goods is not accurate enough and the sizes of the goods are different, resulting in the size of the goods preventing the unmanned forklift from passing through a certain height-limited or other restricted passage or obstacle. The existing technology focuses on the planning of obstacle avoidance routes for unmanned forklifts. Through the intelligent scheduling and path optimization of unmanned forklifts, the driving distance, time, energy consumption and other indicators of the unmanned forklift are minimized while meeting the requirements of cargo transportation. However, the existing technology lacks the planning of unmanned forklift handling routes for cargo sizes. In particular, there is a lack of judgment of the collision relationship between cargo and obstacles for different cargo sizes and postures to adjust the handling route and cargo posture. Simple ranging sensors cannot obtain relevant features of cargo for collision inspection.

[0004] The patent application with publication number CN110872080A discloses a route planning system based on an unmanned forklift, including an unmanned forklift and a monitoring terminal for monitoring the unmanned forklift. The unmanned forklift also includes a motion module and a monitoring module; the motion module also includes a start-stop device for starting and stopping the forklift, a moving object acquisition and capture module for capturing moving objects on the forklift's driving path, and autonomous navigation for path navigation. The unmanned forklift of this patent can optimize the operating route, and capture moving images in real time on the cargo transportation path according to the moving objects on the path to avoid collisions. The differential navigation device can adjust the path in time to ensure the safety of the forklift during operation. The four cameras detect the surroundings of the forklift respectively, and after the images are fed back to the main control device, the captured images are analyzed. After analysis, the vehicle speed can be controlled to ensure the safety of the entire route. The unmanned forklift also has the function of route planning. However, the defect of this patent is that if the length of the cargo carried by the unmanned forklift exceeds the width of the unmanned forklift, the four-bit radar of this patent cannot obtain the distance between the longest end of the cargo and the obstacle, resulting in that the patent ignores the increase in the volume occupied by the unmanned forklift caused by the cargo when planning the route. The cargo may collide with the obstacle, resulting in the unmanned forklift being unable to transport the cargo through the route, and even causing the cargo to fall over or be damaged.

[0005] CN112978619A discloses a shock-absorbing forklift for cargo transportation with an automatic guiding device, including a forklift body, an automatic guiding device, a loading and unloading fork, a loading and unloading fork spacing adjustment device, a forklift lifting frame and a Mecanum wheel set, wherein the automatic guiding device is arranged on the forklift body. The forklift body can travel along the set route without deviation, and the accuracy is high. A Mecanum wheel set is arranged at the bottom of the forklift body, which can reduce the vibration generated during cargo transportation, avoid displacement of cargo during transportation due to vibration, or even fall, and maintain cargo transportation safety; at the same time, the width between the two loading and unloading forks on the forklift body can be adjusted, which can be adjusted according to the actual cargo to be loaded and unloaded, so as to meet the loading and unloading requirements of different types of cargo. This patent takes into account the loading and unloading requirements of different types of cargo, but also does not take into account the problem of different sizes caused by different types of cargo. If the cargo is a long strip, the cargo may be displaced and fall due to the collision between the cargo and the obstacle during transportation, which reduces the efficiency of cargo transportation by unmanned forklift.

[0006] In addition, on the one hand, there are differences in understanding among those skilled in the art; on the other hand, the applicant studied a large number of documents and patents when making the present invention, but due to space limitations, not all details and contents are listed in detail. However, this does not mean that the present invention does not have the characteristics of these prior arts. On the contrary, the present invention already has all the characteristics of the prior art, and the applicant reserves the right to add relevant prior art to the background technology. Summary of the invention

[0007] The existing technology uses sensors to obtain environmental data of freight warehouses so that unmanned forklifts can avoid obstacles and reach the target point smoothly during their driving process. However, the existing technology does not take into account the collision between unmanned forklifts and obstacles caused by the different sizes of goods or incorrect placement of goods when they are being transported. Traditional ranging sensors can only obtain the distance between the unmanned forklift and the obstacle. If the goods are long strips, the goods may collide with the obstacle when the ranging sensor determines that they are far away from the obstacle. In particular, when unmanned forklifts enter height-restricted or curved passages, how to plan suitable unmanned forklift transportation routes based on the size of the goods and the different placement positions is a technical problem that the existing technology urgently needs to solve.

[0008] In order to judge the passability of cargo during the handling process, the prior art has already presented a technical solution for simulating the transport assessment of cargo transport based on a three-dimensional channel model. For example, the patent document with the publication number CN114387407A discloses a method for assessing the passability of large cargo transportation based on a three-dimensional channel model. First, the transportation route is determined, a large-scale transportation vehicle model, a large-scale cargo model, a special road scene model, and an obstacle model are created, and each model is rendered on a map. Then, a three-dimensional channel model is generated, and the positional relationship between the large-scale transportation vehicle model and the large-scale cargo model in the three-dimensional channel model is obtained to determine whether there is contact and collision with the three-dimensional channel model and the obstacle model during transportation. This technical solution uses three-dimensional modeling and simulation technology to simulate the actual transportation process in the software, and through data matching and algorithm analysis, the passability and safety of the transport vehicle are determined. However, this technical solution fixes the perspective on the large-cargo transport vehicle model, simulates the large-cargo transport vehicle model moving along the planned transport route, and obtains the positional relationship between the large-cargo transport vehicle model and the large-cargo model in the three-dimensional channel model. This technical solution assumes that large-cargo has a certain occupied space and that the occupied space will not change. Therefore, the collision analysis at this time cannot take into account the impact of the size change of the cargo on the collision during transportation. This technical solution is suitable for collision analysis during large-scale road traffic, but cannot achieve accurate collision analysis in places such as cargo storage warehouses where there are multiple cargo transportation sizes required.

[0009] In view of the shortcomings of the prior art, according to one aspect of the present invention, a cargo handling route planning method is disclosed, the method comprising: constructing several bounding boxes in a virtual scene of unmanned forklift transportation based on obstacle information; constructing at least one bounding box based on the model data of the unmanned forklift and the size data of the transported cargo; performing a pre-collision on at least one handling route based on several bounding boxes and at least one bounding box to determine whether there is a transportation risk, and in the case of a transportation risk, performing a collision check between the bounding box and the bounding box to obtain three-dimensional collision data; planning the posture of the cargo carried by the unmanned forklift and / or the handling route of the unmanned forklift based on the three-dimensional collision data. Compared with the above-mentioned prior art, the cargo handling route planning method of the present invention can adjust the posture of the cargo carried by the unmanned forklift and / or the handling route of the unmanned forklift according to the results of the collision analysis. Based on the above-mentioned distinguishing technical features, the problems to be solved by the present invention may include: how to adjust the corresponding handling strategy according to the different handled cargoes and reduce the data processing amount of the collision analysis. Specifically, the present invention uses the change in the overall size of the unmanned forklift after carrying the goods to perform a simple pre-collision with several obstacles in the virtual scene of the transportation, thereby using a small amount of calculation to determine whether the unmanned forklift will collide with obstacles on the preset path after carrying the goods, and screen out sections where collisions will not occur. The present invention also performs high-precision collision checks on sections with transportation risks, obtains three-dimensional collision data of collisions, and adjusts the position of the unmanned forklift carrying the goods or the unmanned forklift's transportation route to avoid the section with transportation risks. As a result, not only the amount of preliminary calculations for the collision prediction of the preset path is reduced, but also a safe transportation route for the unmanned forklift can be planned.

[0010] In order to ensure the smoothness and safety of goods during transportation, the prior art has already appeared a technical solution for adjusting the position of goods to achieve orderly transportation of goods in places with large logistics turnover. For example, the patent document with publication number CN115818192A discloses a transportation device that can calibrate the position of goods, which includes a transmission device for driving the goods from the loading position to the unloading position, a positioning device for adjusting the goods in a standing posture to the goods in a lying posture, and a calibration device for adjusting the position of the goods so that it can move on the transmission device in the same posture; when the transmission path of the transmission device is a straight line, the calibration device is configured as two groups, and the upstream end and downstream end of the transmission device are respectively provided; when the transmission path includes at least one turning point, the calibration device is configured as multiple groups, and the calibration device is respectively provided at the upstream end and downstream end of the turning point. This technical solution uses the setting based on the positioning device to adjust the goods from the standing state to the lying state, thereby lowering the center of gravity of the goods, and can be transported to the destination in a good state. The transportation route in this process is fixed and cannot be adjusted according to the installation state of the goods. In other words, the technical solution is limited to adjusting the posture of the cargo to its installation state. The main control factor of the process adjustment is the initial installation state of the cargo. The technical solution does not involve the collision process that may be caused by the external dimensions of the cargo during transportation. In addition, when it is predicted that a collision will occur during the handling process, the installation posture of the cargo on the unmanned forklift and / or the handling route of the unmanned forklift cannot be adjusted.

[0011] According to a preferred embodiment, the method for constructing several bounding boxes includes: establishing a global coordinate system in a virtual scene based on the obstacle information of the freight warehouse that has completed the digital twin; establishing a geometric body corresponding to the obstacle in the global coordinate system based on the longest side of the obstacle in the obstacle information, and using the geometric body as a bounding box. Compared with the above-mentioned prior art, the present invention can establish a collision analysis area for an unmanned forklift in a global coordinate system. Based on the above-mentioned distinguishing technical features, the problem to be solved by the present invention may include: how to reduce the data processing volume of the collision analysis during the transportation process of the unmanned forklift. Specifically, due to the different specific goods stored in the freight warehouse, the size or limited range of the space area occupied by the shelves for placing various goods is also different. At the same time, the different structures such as inlet and outlet channels set on different shelves also directly affect their specific appearance structure. To realize the handling process of different goods between different shelves, it is necessary to perform the collision analysis process according to different shelf structure forms and goods types, which leads to a large initial collision analysis data processing volume. On the contrary, in the present invention, the constructed bounding box is not a bounding box that is completely fitted or conformed to the obstacle, but a regular geometric body that can approximately characterize the range covered by the obstacle. This avoids the problem that the bounding box established in the prior art needs to accurately calculate every change and shape of the obstacle, simplifies the calculation amount of subsequent collision checks, and at the same time does not miss several obstacles in the freight warehouse.

[0012] According to a preferred embodiment, the method for constructing a bounding box includes: establishing a virtual model for characterizing the body size of the unmanned forklift based on the model data of the unmanned forklift; when the unmanned forklift carries cargo, constructing at least one bounding box containing the unmanned forklift and the cargo based on the virtual model and the posture and size data of the cargo. Preferably, when the posture of the cargo changes, the bounding box is updated or rebuilt based on the changed posture of the cargo. When the posture of the cargo changes, the processing unit updates or rebuilt the bounding box based on the changed posture of the cargo. The bounding box refers to the smallest cube containing the unmanned forklift and the cargo. Thus, the present invention can realize the planning of the handling route of the unmanned forklift through collision checking of the bounding box and the bounding box.

[0013] According to a preferred embodiment, the method further includes: establishing a virtual channel that can characterize the motion trajectory of the unmanned forklift by moving the bounding box along a preset transport route; cutting the virtual channel based on the three-dimensional sections of several bounding boxes to form several bounding boxes to be tested. The shape of the virtual channel is curved and complex. If a global collision check is performed on the virtual channel, not only will the calculation amount of the processing unit be increased, but most of the collision calculations are also meaningless. Therefore, the present invention is based on the bounding box, and selects the bounding box to be tested corresponding to the unmanned forklift in the area where the bounding box exists to perform a collision check, thereby improving the efficiency of data calculation.

[0014] According to a preferred embodiment, the pre-collision method includes: judging the collision relationship between the bounding box to be tested and the corresponding bounding box to screen out several bounding boxes to be tested and several bounding boxes that have a collision relationship. The processing unit determines whether there is a collision relationship between the bounding box to be tested and the bounding box through spatial judgment in the global coordinate system, and then screens out several bounding boxes to be tested and several bounding boxes that may have a collision relationship. The present invention screens out obstacles that may collide with the unmanned forklift through a pre-collision of a larger range of bounding boxes, and then divides the bounding boxes with high precision for collision checking.

[0015] According to a preferred embodiment, the method further includes: when there is a collision relationship between the bounding box to be tested and the bounding box, the irregular bounding box is divided into several bounding box sub-modules in a segmented and / or split manner, so as to screen out the bounding box sub-modules that may have a collision relationship by pre-colliding with the bounding box to be tested through several bounding box sub-modules. Compared with the above-mentioned prior art, the present invention can divide the part where the bounding box to be tested and the bounding box have a collision relationship. Based on the above-mentioned distinguishing technical features, the problem to be solved by the present invention may include: how to screen some bounding boxes that have a collision risk to improve the accuracy of collision analysis. Specifically, since the space covered by the bounding box constructed according to the longest side of the obstacle exceeds the actual volume of the obstacle, the accuracy of collision relationship judgment is too low. Therefore, after screening out some bounding boxes, the present invention performs a secondary screening by dividing the bounding box into multiple bounding box sub-modules, thereby increasing the accuracy of collision relationship judgment on the basis of reducing the amount of calculation of three-dimensional collision data.

[0016] According to a preferred embodiment, the method for obtaining three-dimensional collision data includes: after screening out a number of bounding boxes and a number of bounding box submodules, dividing the number of bounding boxes and the number of bounding box submodules into a number of boundary elements to obtain three-dimensional collision data by performing at least one collision check with a bounding box. The present invention performs a point-by-point collision check with the bounding box on each boundary element to obtain three-dimensional collision data of the collision between the number of bounding boxes and the number of bounding box submodules and the bounding box. Thus, the collision range of the bounding box and the bounding box can be accurately obtained by calculation, and some lines that do not have a collision relationship are retained, and some lines that have a collision relationship are adjusted based on the three-dimensional collision data.

[0017] According to a preferred embodiment, the method further includes: controlling the unmanned forklift to change the posture of the goods in a translation or rotation manner based on the three-dimensional collision data so that the bounding box does not collide with the bounding box; after changing the posture of the goods, if there is still a collision relationship between the bounding box and the bounding box, a transportation route that the unmanned forklift can safely pass is planned based on a combination of several bounding boxes that do not have a collision relationship. Compared with the above-mentioned prior art, the present invention can avoid collision by adjusting the posture of the goods, and can adjust the transportation route according to the change of the posture of the goods. Based on the above-mentioned distinguishing technical features, the problems to be solved by the present invention may include: how to combine and plan the transportation route according to the type and status information of the goods transported by the unmanned forklift. Specifically, in large-scale transportation processes, it is unnecessary to re-plan the transportation route when encountering obstacles. In some cases, the goods can be passed through a certain passage or obstacle by changing the posture of the goods. For example, by tilting the goods to pass through a certain height-limited passage. The present invention enables the unmanned forklift to pass through a passage or obstacle that could not be passed before by changing the posture of the goods. When encountering a height-restricted passage, the vertical placement of the goods can be adjusted to an inclined or horizontal placement to lower the height of the goods so that they can pass through the height-restricted passage. In the above judgment process, no staff intervention is required, and the intelligent control of the unmanned forklift is achieved with less calculation, which will not affect the transportation efficiency of the unmanned forklift.

[0018] Another aspect of the present invention relates to a cargo handling route planning system, including an unmanned forklift and a processing unit. The processing unit is configured to: construct several bounding boxes in a virtual scene of unmanned forklift transportation based on obstacle information; construct at least one bounding box based on the model data of the unmanned forklift and the size data of the transported cargo; perform pre-collision on at least one handling route based on several bounding boxes and at least one bounding box to determine whether there is a transportation risk, and in the case of a transportation risk, perform a collision check between the bounding box and the bounding box to obtain three-dimensional collision data; and plan the posture of the cargo carried by the unmanned forklift and / or the handling route of the unmanned forklift based on the three-dimensional collision data. The present invention establishes a corresponding bounding box through the change of the collision volume of the unmanned forklift after carrying the cargo, so that the unmanned forklift can avoid collision with the cargo while avoiding obstacles. The processing unit of the present invention can plan the handling route of the unmanned forklift in the case of carrying cargo based on the formed bounding box, effectively avoiding the unmanned forklift from being able to safely reach the target location along the pre-set handling route in the case of carrying due to changes in the size or posture of the cargo, and preventing possible collisions between the cargo and obstacles while ensuring the transportation efficiency of the unmanned forklift.

[0019] According to a preferred embodiment, the processing unit is further configured to: control the unmanned forklift to change the posture of the goods in a translation or rotation manner based on the three-dimensional collision data so that the bounding box does not collide with the bounding box; after changing the posture of the goods, if there is still a collision relationship between the bounding box and the bounding box, plan a transportation route that the unmanned forklift can safely pass based on a combination of several bounding boxes that do not have a collision relationship. The processing unit of the present invention adjusts the posture of the goods by controlling the unmanned forklift to clamp and rotate the goods, so that the unmanned forklift can pass through a certain height-limited, width-limited or curved passage and obstacle. This method does not require the processing unit to re-plan the transportation route. The relevant three-dimensional collision data obtained by the collision check can realize the adjustment of the posture of the goods, thereby avoiding the possible collision of the goods with the passage or obstacles along the preset transportation route, thereby improving the efficiency of the unmanned forklift in transporting goods. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 This is a simplified module connection relationship diagram of a cargo handling route planning system according to a preferred embodiment of the present invention;

[0021] Figure 2 It is a simplified flow chart of a cargo handling route planning method according to a preferred embodiment of the present invention;

[0022] Figure 3 This is a simplified application scenario diagram of a cargo handling route planning system according to a preferred embodiment of the present invention;

[0023] Figure 4 It is a simplified application scenario schematic diagram of a collision between an unmanned forklift and an obstacle in a cargo handling route planning system according to a preferred embodiment of the present invention;

[0024] Figure 5 This is another simplified application scenario schematic diagram of a cargo handling route planning system according to a preferred embodiment of the present invention in which an unmanned forklift collides with an obstacle;

[0025] Figure 6 It is a simplified application scenario schematic diagram of a cargo handling route planning system according to a preferred embodiment of the present invention, in which an unmanned forklift adjusts the cargo posture so as not to collide with obstacles;

[0026] Figure 7 It is another simplified application scenario schematic diagram of a cargo handling route planning system according to a preferred embodiment of the present invention, in which an unmanned forklift adjusts the cargo position so as not to collide with obstacles.

[0027] Reference numerals list

[0028] 100: processing unit; 200: unmanned forklift; 201: bounding box; 300: obstacle; 301: bounding box; 302: bounding box submodule; 303: height-limited channel; 304: shelf; 400: goods; 500: transportation route. DETAILED DESCRIPTION

[0029] The following is a detailed description with reference to the accompanying drawings.

[0030] Unmanned forklift 200: Unmanned forklift 200 is an intelligent industrial vehicle robot that can automatically complete various handling and transportation tasks without manual driving through various navigation technologies. In the present invention, the unmanned forklift 200 can be a clamping and rotating intelligent forklift, thereby being able to control the translation and rotation of the cargo 400.

[0031] Bounding box 301: refers to a geometric body that can completely contain an object or a group of objects. Bounding box 301 can be two-dimensional or three-dimensional, and can be axis-aligned or rotated. In the present invention, bounding box 301 refers to a cube constructed with the longest side length of obstacle 300, which not only contains obstacle 300, but also contains additional space.

[0032] Bounding box 201: refers to a minimum cube that can completely contain an object or a group of objects. Bounding box 201 is usually axis-aligned, that is, its sides are parallel to the coordinate axes.

[0033] Collision check: Determine whether there is a spatial overlap between at least two virtual models in the virtual space.

[0034] Virtual channel: a volume formed by the bounding box 201 during the movement. For example, in mathematics and geometry, a volume is formed by the movement of a surface, that is, a plane figure moves along a path to form a three-dimensional figure. In the present invention, the virtual channel refers to a three-dimensional figure formed by the bounding box 201 during the movement along the transport route 500.

[0035] Position and posture: In the present invention, the position and posture represent the position and posture of the cargo 400. Any rigid body (such as the cargo 400) can be accurately and uniquely represented by its position and posture in the global coordinate system.

[0036] Global coordinate system: refers to the coordinate system where objects are located in the three-dimensional space of the virtual scene. In the present invention, the vertex coordinates of each object (such as obstacle 300, unmanned forklift 200, and cargo 400, etc.) can be expressed based on this coordinate system. The global coordinate system is the coordinate system of the entire virtual scene and is a fixed coordinate system. Each object is displaced, rotated, and scaled relative to this coordinate system.

[0037] Example 1

[0038] In the prior art, the working process of the intelligent forklift does not require human control. However, in the process of transporting goods, the intelligent forklift will inevitably encounter some unpredictable obstacles 300, and the forklift cannot travel along the originally planned path due to the influence of the obstacles 300. In particular, when the goods 400 carried by the intelligent forklift are of different sizes, it is difficult for the intelligent forklift to determine whether the pre-planned path is affected by the change in the size of the goods 400, causing it to collide with the obstacle 300 on the path. For example, if the goods 400 carried by the intelligent forklift are large in size, it may not be suitable for the pre-planned path, and the goods 400 need to be transported through other paths or channels. At present, the prior art mostly uses artificial potential field method, bubble belt technology, etc. to perform obstacle avoidance planning for the intelligent forklift, which avoids obstacles by planning the obstacles 300 on the path as larger bounding boxes 201. However, due to the problem of calculation accuracy, this method easily causes too many or too large bounding boxes 201 on the path, which makes the intelligent forklift appear to have no suitable moving path. In addition, the path planned by this method not only fails to consider the size of the goods 400 and the corresponding placement posture, but also fails to consider the excessive redundant data and excessive calculation amount caused by calculating the bounding box 201 for each obstacle 300 on the path, resulting in the problem of insensitivity of the intelligent forklift and low transportation efficiency. Therefore, how to plan a suitable forklift moving path based on the size and placement posture of the goods 400 and how to reduce the calculation amount of obstacle 300 calculation and path planning are technical problems to be solved by the present invention.

[0039] like Figure 1 As shown, the present invention discloses a planning system for how to realize rapid dispatch of goods 400 in a freight warehouse by changing the posture of the transported goods 400 when the unmanned forklift 200 is carrying the goods 400, or planning the transportation route 500 of the unmanned forklift 200 when carrying the goods 400 in a scenario where an unmanned forklift 200 is required to transport the goods 400.

[0040] To this end, the cargo transport route 500 planning system of the present invention includes at least one unmanned forklift 200 that carries cargo 400 by clamping and rotating the cargo and a processing unit 100 for planning the transport route 500 of the unmanned forklift 200 and the corresponding posture of the cargo 400.

[0041] When the processing unit 100 simulates several bounding boxes 301 representing obstacles 300 in the virtual scene, or when the freight warehouse that has completed the digital twin needs to transport goods 400, several unmanned forklifts 200 parked at the charging station communicate with the processing unit 100 in the dispatch room or monitoring room of the remote server or freight warehouse in a wired or wireless manner. Several unmanned forklifts 200 are docked in the charging station in a wired connection to replenish energy, and several unmanned forklifts 200 are simultaneously connected to the processing unit 100 in a wired connection in the charging station to transmit the transport route 500 simulated by the processing unit 100 in the virtual scene and / or the planning of the posture of the unmanned forklift 200 to transport the goods 400 corresponding to the transport route 500.

[0042] In the present invention, the processing unit 100 can be arranged in the dispatching room or monitoring room of the freight warehouse or in the charging station where the unmanned forklift 200 is parked. When the processing unit 100 is arranged in the charging station, the processing unit 100 establishes a communication connection with several unmanned forklifts 200 in a wired manner. The charging station is provided with a communication interface connected to the processing unit 100. The processing unit 100 sends the transport route 500 and the posture of the goods 400 simulated in the virtual scene to the unmanned forklift 200 in a wired manner according to the goods to be transported by the unmanned forklift 200. When the unmanned forklift 200 is required to transport the goods, due to the large amount of relevant data of the transport route 500 and the corresponding posture of the goods 400, it is difficult to quickly transmit the data by wireless means, and a large bandwidth support is required. Therefore, the wired transmission method is more suitable for the transmission of the transport route 500 obtained by the processing unit 100 after pre-simulation than the wireless transmission method. In this case, the processing unit 100 pre-simulates and obtains the transport route 500 of the unmanned forklift 200 and the corresponding position of the cargo 400 in the virtual scene, and transmits some data in a wired manner during the charging process of the unmanned forklift 200.

[0043] Figure 3 to Figure 7 The three-dimensional simulated virtual scene of the freight warehouse. When the processing unit 100 constructs a plurality of bounding boxes 301 in the virtual scene transported by the unmanned forklift 200 based on the obstacle information, and the processing unit 100 constructs the bounding box 201 based on the model data of the unmanned forklift 200 and the size data of the transported goods 400, the unmanned forklift 200 carrying the goods 400 moves along the transport route 500 received from the processing unit 100. Since the transport route 500 transmitted by the processing unit 100 is accompanied by the posture of the goods 400 to be transported when passing through the transport route 500, when the unmanned forklift 200 approaches a certain obstacle 300, it passes through the obstacle 300 by adjusting the posture of the transported goods 400.

[0044] like Figure 3As shown, the unmanned forklift 200 needs to pass through a height-limited passage 303 or needs to pass through a fire curtain that has been lowered. Figure 4 As shown, due to the center of gravity requirement of the unmanned forklift 200 for carrying the goods 400, the height of the goods 400 carried by the unmanned forklift 200 is relatively high, which makes it impossible for the unmanned forklift 200 to pass through the height-limited channel 303 with the current position of the goods 400. The unmanned forklift 200 can pass through the height-limited channel 303 by detouring, but this affects the transportation efficiency of the unmanned forklift 200. In the present invention, due to the pre-planning of the processing unit 100 in the virtual scene, the unmanned forklift 200 can pass through the height-limited channel 303 by changing the position of the goods when approaching the height-limited channel 303. The processing unit 100 performs a pre-collision on at least one transport route 500 based on a plurality of bounding boxes 301 and bounding boxes 201 to determine whether there is a transportation risk.

[0045] like Figure 5 As shown, in the case of transportation risk, the processing unit 100 performs a collision check between the bounding box 301 and the bounding box 201 to obtain three-dimensional collision data. The processing unit 100 plans the posture of the cargo 400 carried by the unmanned forklift 200 and / or the transport route 500 of the unmanned forklift 200 based on the three-dimensional collision data. The above calculations are all obtained by simulation in the virtual scene performed by the processing unit 100, and in particular, the processing unit 100 can send the transport route 500 and the corresponding cargo posture to the unmanned forklift 200 by wired transmission. Figure 6 and Figure 7 As shown, when the unmanned forklift 200 approaches the height-limited passage 303 , it passes through the height-limited passage 303 in a manner of lowering the position of the cargo 400 .

[0046] The unmanned forklift 200 of the present invention can intelligently transport the cargo 400 at the cargo storage point or cargo storage warehouse. The cargo storage point or cargo storage warehouse has completed the digital twin, so that the processing unit 100 can obtain the obstacle information in the cargo storage point or cargo storage warehouse. The processing unit 100 of the present invention can be one or more of a processor, a server, a cloud platform, a computer, and an intelligent device, which is used to execute the software program of the cargo handling route 500 planning method of the present invention.

[0047] The present invention proposes a cargo transport route 500 planning method, such as Figure 2 As shown, the method includes:

[0048] S1: The processing unit 100 constructs a number of bounding boxes 301 in the virtual scene transported by the unmanned forklift 200 based on obstacle information.

[0049] The above-mentioned bounding box 301 refers to a rectangle or a cube that can completely contain an object or a group of objects. In the present invention, the bounding box 301 refers to a cube that can completely contain the obstacle 300 on the transport path of the unmanned forklift 200.

[0050] S2: The processing unit 100 constructs at least one bounding box 201 based on the model data of the unmanned forklift 200 and the size data of the transported cargo 400 .

[0051] S3: The processing unit 100 performs a pre-collision on at least one transport route 500 based on several bounding boxes 301 and at least one bounding box 201 to determine whether there is a transportation risk. If there is a transportation risk, the bounding box 301 is subjected to a collision check with the bounding box 201 to obtain three-dimensional collision data of the collision between the bounding box 301 and the bounding box 201.

[0052] S4: The processing unit 100 plans the position of the cargo 400 carried by the unmanned forklift 200 and / or the transport route 500 of the unmanned forklift 200 based on the three-dimensional collision data.

[0053] The present invention uses the change in the overall size of the unmanned forklift 200 after carrying the cargo 400 to perform a simple pre-collision with several obstacles 300 in the virtual scene of the transportation, thereby using a small amount of calculation to determine whether the unmanned forklift 200 collides with the obstacle 300 on the preset path after carrying the cargo 400, and screen out the section where the collision will not occur. The processing unit 100 of the present invention also performs a high-precision collision check on the section with transportation risks, obtains the three-dimensional collision data of the collision, and adjusts the position of the unmanned forklift 200 carrying the cargo 400 or the way the unmanned forklift 200 transports the route 500 to avoid the section with transportation risks. As a result, not only the amount of preliminary calculations for the collision prediction of the preset path is reduced, but also a safe transportation route 500 for the unmanned forklift 200 can be planned.

[0054] S11: The processing unit 100 establishes a global coordinate system in a virtual scene based on the obstacle information of the freight warehouse for which the digital twin has been completed.

[0055] A number of digital twin sensors are provided in the freight warehouse for which digital twins have been completed. The digital twin sensor can be a laser scanning sensor, thereby obtaining global obstacle information of the current freight warehouse and the relative coordinates of the obstacles relative to each digital twin sensor. The processing unit 100 converts the relative coordinates of the obstacles 300 into global coordinates in the global coordinate system based on the obstacle information and the global coordinates of the current position points of each digital twin sensor, thereby establishing the obstacles 300 in the virtual scene.

[0056] S12 : The processing unit 100 establishes a number of bounding boxes 301 corresponding to the obstacles 300 based on the global coordinate system and the sizes of the obstacles 300 .

[0057] Preferably, step S12: the processing unit 100 establishes a geometric body corresponding to the obstacle 300 based on the longest side of the size of the obstacle 300, and uses the geometric body as the bounding box 301. The processing unit 100 uses the longest side length of the obstacle 300 as the boundary of the bounding box 301, thereby constructing the bounding box 301 that completely contains the obstacle 300 in the virtual scene.

[0058] like Figure 1 As shown, the above-mentioned obstacles 300 may refer to several shelves 304 in the freight warehouse, parked unmanned forklifts 200 or other stationary obstacles 300. In the present invention, the bounding box 301 constructed by the processing unit 100 is not a bounding box 201 that completely fits or conforms to the obstacle 300, but a regular geometric body that can approximately characterize the range covered by the obstacle 300. Therefore, the problem that the bounding box 201 established in the prior art needs to accurately calculate every change and shape of the obstacle 300 is avoided, the calculation amount of the subsequent collision check is simplified, and at the same time, several obstacles 300 in the freight warehouse will not be omitted.

[0059] S21: The processing unit 100 establishes a virtual model for representing the body size of the unmanned forklift 200 based on the model data of the unmanned forklift 200 .

[0060] The model data of the unmanned forklift 200 can be pre-stored in the database according to its different models. The processing unit 100 wirelessly connected to the unmanned forklift 200 can identify the model of the current unmanned forklift 200, and call the model data corresponding to the model from the database to build a virtual model representing the body size of the unmanned forklift 200. The processing unit 100 can also pre-build a virtual model corresponding to the unmanned forklift 200 to directly call the virtual model to plan the transport route 500. In the present invention, the unmanned forklift 200 can be a clamping rotary forklift or other intelligent forklift that can realize the rotation and placement of the goods 400, thereby being able to pick up the goods 400 on the shelf 304 and adjust the posture of the goods 400 during the transportation of the goods 400.

[0061] S22 : The processing unit 100 constructs at least one bounding box 201 based on the size data and position of the cargo 400 carried by the unmanned forklift 200 and the virtual model of the unmanned forklift 200 .

[0062] The cargo 400 will form bounding boxes 201 of different sizes and shapes due to different placement methods on the unmanned forklift 200. For example, the unmanned forklift 200 can place the cargo 400 on the cantilever or fork in a horizontal or vertical manner. Preferably, the unmanned forklift 200 is provided with a detection mechanism for monitoring the size of the cargo 400 carried. The cantilever or fork of the unmanned forklift 200 is provided with a detection mechanism, which can be a photoelectric or distance sensor, so as to obtain the position and size of the cargo 400 carried by the unmanned forklift 200, and the position and size data can be transmitted to the processing unit 100. Since the unmanned forklift 200 in the present invention is an intelligent forklift, it can adjust the position and size of the cargo 400. Preferably, when the goods 400 on the shelf 304 are not aligned with the forks, the unmanned forklift 200 can adjust the angle of the forks to pick up the goods 400 on the shelf 304, and the unmanned forklift 200 uses its rotating clamp-type mechanical claw to adjust the position of the goods 400 on the forks to reduce the overall space occupied by the unmanned forklift 200 and the carried goods 400.

[0063] Preferably, when the processing unit 100 obtains the position and size of the cargo 400 carried by the unmanned forklift 200, the processing unit 100 establishes at least one bounding box 201 of the unmanned forklift 200 carrying the cargo 400 in combination with the virtual model of the unmanned forklift 200. When the position of the cargo 400 changes, the processing unit 100 updates or rebuilds the bounding box 201 based on the changed position of the cargo 400. The bounding box 201 refers to the smallest cube containing the unmanned forklift 200 and the cargo 400. Therefore, the present invention can realize the planning of the transport route 500 of the unmanned forklift 200 through the collision check of the bounding box 201 and the bounding box 301.

[0064] S31: The processing unit 100 moves the bounding box 201 along the preset transport route 500 to establish a virtual channel that can represent the motion trajectory of the unmanned forklift 200.

[0065] The virtual channel refers to a three-dimensional channel formed after the bounding box 201 that can represent the unmanned forklift 200 and the transported goods 400 moves on the preset transport route 500. The processing unit 100 records each trail during the movement of the bounding box 201 in a delayed manner, thereby forming a three-dimensional virtual channel. For example, when the bounding box 201 is spherical, the bounding box 201 moves along the straight preset transport route 500 to form a cylinder-like virtual channel. The virtual channel of the present invention can represent the movement trajectory of the bounding box 201 (the goods 400 and the unmanned forklift 200 carrying it) during the movement along the transport route 500 and the space it contacts. The present invention determines whether the unmanned forklift 200 collides with the obstacle 300 under the transport route 500 through the pre-collision of the virtual channel and the bounding box 301, thereby planning a transport route 500 suitable for the size of the goods 400 and the unmanned forklift 200. Since the transport route 500 is not a single straight line in most cases, the virtual channel formed by the present invention is a channel that bends and extends in the virtual scene.

[0066] S32: The processing unit 100 cuts virtual channels based on the three-dimensional sections of the plurality of bounding boxes 301 to form a plurality of bounding boxes to be measured.

[0067] Specifically, since most of the obstacles 300 in the freight warehouse are regular geometric bodies (such as shelves 304), the processing unit 100 creates various sections tangent to the bounding box 301 along at least three mutually perpendicular directions of the global coordinate system based on the established global coordinate system. In the case of an extended virtual channel, the processing unit 100 divides the virtual channel into bounding boxes to be tested based on a number of bounding boxes 301 and retains them in the area of ​​the bounding box 301. The shape of the virtual channel is curved and complex. If a global collision check is performed on the virtual channel, not only will the calculation amount of the processing unit 100 be increased, but most of the collision calculations are also meaningless. Therefore, based on the bounding box 301, the present invention screens out the bounding box to be tested corresponding to the unmanned forklift 200 in the area where the bounding box 301 exists to perform a collision check, thereby improving the efficiency of data calculation.

[0068] S33: the processing unit 100 determines the collision relationship between the bounding box to be tested and the bounding box 301 based on the pre-collision between the bounding box to be tested and the corresponding bounding box 301 to screen out several bounding boxes to be tested and several bounding boxes 301 having a collision relationship.

[0069] The above-mentioned collision relationship includes intersection or inclusion. The collision relationship means that the bounding box to be tested and the bounding box 301 overlap in three-dimensional space, and there may be a risk of collision between the two. The processing unit 100 determines whether there is a collision relationship between the bounding box to be tested and the bounding box 301 through spatial judgment in the global coordinate system, and then screens out several bounding boxes to be tested and several bounding boxes 301 that may have a collision relationship.

[0070] Since the bounding box 301 constructed by the present invention is formed according to the longest side of the obstacle 300. If the obstacle 300 is a regular geometric body, the pre-collision between the bounding box 301 and the virtual channel is accurate. If the obstacle 300 is an irregular solid, the bounding box 301 includes the real position range of the obstacle 300. At this time, there must be a large error in judging whether the unmanned forklift 200 carrying the cargo 400 collides with the obstacle 300. In the present invention, the acquisition of three-dimensional collision data is cumbersome and requires a large amount of calculation. Therefore, it should be further determined whether the bounding box 301 has a collision relationship with the bounding box 201. Although the bounding box 301 simplifies the calculation amount of obtaining the virtual model of the obstacle 300 in the early stage, it also expands the volume of the obtained virtual model, resulting in errors in the subsequent judgment of the collision relationship.

[0071] It should be noted that the reason why the bounding box 301 is not constructed based on each real contour of the obstacle 300 is that the bounding box 301 of the obstacle 300 constructed by the real contour needs to consume a large amount of computing resources, and the amount of data calculation required to construct the bounding box 301 of each real contour cannot be ignored. The obstacles 300 that may collide with the unmanned forklift 200 are not the majority. Therefore, the present invention screens out the obstacles 300 that may collide with the unmanned forklift 200 through the pre-collision of the bounding box 301 with a larger range, and then divides the bounding box 301 with high precision for collision detection.

[0072] S34: When there is a collision relationship between the bounding box to be measured and the bounding box 301, the processing unit 100 divides the irregular bounding box 301 into several bounding box sub-modules 302 in a segmented and / or split manner, so as to screen out the bounding box sub-modules 302 that may have a collision relationship by pre-colliding the several bounding box sub-modules 302 with the bounding box to be measured.

[0073] Since the space covered by the bounding box 301 constructed according to the longest side of the obstacle 300 exceeds the actual volume of the obstacle 300, the accuracy of the collision relationship judgment is too low. Therefore, after filtering out part of the bounding box 301, the present invention performs a secondary screening by dividing the bounding box 301 into a plurality of bounding box submodules 302, thereby reducing the amount of three-dimensional collision data calculation and increasing the accuracy of collision relationship judgment.

[0074] Specifically, the processing unit 100 connects the vertices at both ends of the irregular obstacle 300 to construct a minimum virtual boundary that can locally characterize the obstacle 300. The processing unit 100 pre-collides the virtual boundary with the bounding box to be tested to screen out the bounding box 301 with a collision relationship. Preferably, the processing unit 100 segments the bounding box 301 that has undergone secondary screening based on irregular features as a bounding box submodule 302. The above-mentioned irregular features refer to the bending or deformation of the obstacle 300. The judgment of the above-mentioned virtual boundary and the bounding box to be tested can obtain the relative position of the bounding box to be tested and the bounding box 301, and exclude the bounding box to be tested that is far away from the actual obstacle 300. The above-mentioned simple division method can reduce the amount of calculation when constructing the bounding box submodule 302, and the division and judgment process can be completed by simple global coordinate judgment.

[0075] For example, if the obstacle 300 is a bent shelf. When it is constructed as a bounding box 301, the bounding box 301 includes a large number of non-obstacles 300. In this regard, the present invention connects the diagonal vertices of the obstacle 300 to construct a smaller virtual boundary that can contain the obstacle 300. The processing unit 100 pre-collides the virtual boundary with the bounding box to be tested, thereby further determining the collision relationship between the bounding box to be tested and the obstacle 300, and screening out the bounding box 301 with a collision relationship. The processing unit 100 segments the bounding box 301 that has undergone secondary screening based on the irregular features of the obstacle 300, thereby obtaining a number of bounding box sub-modules 302. Based on the pre-collision of a number of bounding box sub-modules 302 with the bounding box to be tested, the processing unit 100 once again excludes the bounding box sub-modules 302 that do not have a collision relationship. The bounding box sub-module 302 thus obtained has a high probability of having a collision relationship with the bounding box to be tested. The present invention uses a pre-collision method to monitor the collision relationship between the bounding box 301 and the bounding box submodule 302 and the bounding box to be tested. It only needs to obtain whether there is a collision relationship, and there is no need to obtain the actual collision intersection, thereby saving a lot of computing resources. Combined with the simple construction method and simple division method of the bounding box 301, the present invention greatly improves the judgment and planning rate of the transportation route 500, and increases the working efficiency of the unmanned forklift 200.

[0076] S35: When a plurality of bounding boxes 301 and a plurality of bounding box submodules 302 are screened out, the processing unit 100 divides the plurality of bounding boxes 301 and the plurality of bounding box submodules 302 into a plurality of boundary elements to perform at least one collision check with the bounding box 201 to obtain three-dimensional collision data.

[0077] The processing unit 100 divides a number of bounding boxes 301 and a number of bounding box submodules 302 into a number of boundary elements in a preset interval unit. The above-mentioned interval unit is, for example, a length unit such as centimeters and millimeters. The present invention performs a point-by-point collision check on each boundary element and the bounding box 201 to obtain three-dimensional collision data of the collision between the bounding box 201 and the bounding box 201. Specifically, the processing unit 100 calculates the global coordinates of the intersection of each boundary element and the bounding box 201, thereby storing the global coordinates of each intersection in a set manner. Several intersections indicate that there is a collision between the bounding box 201 and the boundary element (i.e., the unmanned forklift 200 carrying the cargo 400 and the obstacle 300). In this way, the collision range of the bounding box 301 and the bounding box 201 can be accurately obtained by calculation, and some lines that do not have a collision relationship are retained, and some lines that have a collision relationship are adjusted based on the three-dimensional collision data.

[0078] In the present invention, the above-mentioned reserved partial routes without collision relationship include not only the partial routes without collision relationship of the unmanned forklift 200 carrying the goods 400 on the preset transport route 500, but also the partial routes in the freight warehouse calculated by the processing unit 100 when there is surplus computing power and without collision relationship with the unmanned forklift 200 carrying the goods 400. When the unmanned forklift 200 advances to a certain area, the processing unit 100 selectively reserves the partial routes without collision relationship in the future, so that after the posture of the unmanned forklift 200 carrying the goods 400 changes, the unmanned forklift 200 can be controlled to change the posture of the goods 400 again to return to the preset transport route 500. When the posture of the goods 400 on the unmanned forklift 200 is changed, the processing unit 100 calculates the partial routes without collision relationship with the unmanned forklift 400 based on the changed posture of the goods 400. The processing unit 100 plans at least one transport route 500 based on the continuity of the reserved partial routes without collision relationship, and also includes the posture of the goods 400 required for the unmanned forklift 200 to carry the goods 400 through the partial routes. The above continuity means that the transport route 500 planned by the processing unit 100 is based on continuity as a criterion, rather than the posture of the goods 400 changed by the unmanned forklift 200. When the unmanned forklift 200 arrives at a certain route, the processing unit 100 controls the unmanned forklift 200 to adjust the posture of the goods 400 to pass through the route.

[0079] Thus, the present invention avoids the situation that after the collision transport route 500 is adjusted, the changed position of the cargo 400 or the transport route 500 changes, resulting in collisions at other places, and avoids multiple adjustments to the transport route 500 of the unmanned forklift 200. The present invention preferentially adjusts the position of the cargo 400 carried by the unmanned forklift 200. When the unmanned forklift 200 cannot pass through the current route even after adjusting the position of the cargo 400, the processing unit 100 adjusts the transport route 500 again, and prevents multiple adjustments to the transport route 500 of the unmanned forklift 200.

[0080] Preferably, the processing unit 100 represents the bounding box 201 as a parametric equation and the boundary element as a standard equation, so as to calculate the intersection based on the parametric equation and the standard equation. The parametric equation can be an equation that depicts the bounding box 201 with the parameters of the starting point of the bounding box 201 and the end point of the bounding box 201. The standard equation can be an equation for representing a surface in global coordinates. For example, the parametric equation can be: f(n)=f1+n(f2-f1), where f1 is the starting point of the bounding box 201 and f2 is the end point of the bounding box 201. The standard equation can be: ax+by+cz+d=0, where a, b, c, and d are the coefficients of the boundary element in global coordinates, respectively. Preferably, the processing unit 100 substitutes the parametric equation of the bounding box 201 into the standard equation of the boundary element to calculate the value of n. If 0≤n≤1, it is determined that the bounding box 201 intersects with the boundary element. At this time, the specific intersection coordinates can be obtained by substituting the value of n into the parametric equation. If n<0 or n>1, it is determined that the bounding box 201 does not intersect with the boundary element. The processing unit 100 stores the intersection coordinates in a set as three-dimensional collision data.

[0081] S41: The processing unit 100 adjusts the position and posture of the cargo 400 carried by the unmanned forklift 200 based on the three-dimensional collision data.

[0082] The unmanned forklift 200 of the present invention can carry the cargo 400 by clamping and rotating. As a result, the position of the cargo 400 carried by the unmanned forklift 200 can be adjusted. The sizes of the cargo 400 are different, and the unmanned forklift 200 carries the cargo in different ways. If the unmanned forklift 200 carries the cargo 400 in a horizontal or vertical manner, the size and shape of the bounding box 201 formed by the unmanned forklift 200 will be different, which will result in different positions of the cargo 400 under the preset transport route 500, which will determine whether the unmanned forklift 200 can pass through the obstacle 300.

[0083] Preferably, the processing unit 100 controls the unmanned forklift 200 to change the posture of the cargo 400 by translation or rotation based on the three-dimensional collision data, so that the bounding box 201 does not collide with the bounding box 301. The processing unit 100 determines whether the unmanned forklift 200 can not collide with the obstacle 300 after the cargo 400 is translated or rotated based on the intersection between the bounding box 201 and the bounding box 301 in the three-dimensional collision data. The present invention enables the unmanned forklift 200 to pass through a passage or obstacle 300 that could not be passed before by changing the posture of the cargo 400. In large-scale transportation processes, it is unnecessary to re-plan the transport route 500 when encountering an obstacle 300. In some cases, the cargo 400 can pass through a certain passage or obstacle 300 by changing the posture of the cargo 400. For example, by tilting the cargo 400 to pass through a certain height-limited passage. The obstacle avoidance transportation process in the prior art lacks consideration of the posture of the cargo 400, resulting in the need for the unmanned forklift 200 to replan its transportation route 500 when encountering an "inaccessible" passage or obstacle 300. However, this "inaccessible" does not mean that it is completely inaccessible. The present invention obtains three-dimensional collision data, and on this basis controls the unmanned forklift 200 to adjust the posture of the cargo 400 in a clamping and rotating manner, so that the passage or obstacle 300 encountered by the unmanned forklift 200 can be targeted for obstacle avoidance. In particular, when encountering a height-limited passage, the height of the cargo 400 can be reduced by adjusting the vertical placement of the cargo 400 to an inclined or horizontal placement to pass through the height-limited passage 303. In the above judgment process, there is no need for the intervention of staff, and the intelligent control of the unmanned forklift 200 is achieved through a small amount of calculation, which will not affect the transportation efficiency of the unmanned forklift 200.

[0084] S42: The processing unit 100 plans a transport route 500 of the unmanned forklift 200 based on the three-dimensional collision data.

[0085] In the case where the collision between the unmanned forklift 200 and the obstacle 300 cannot be avoided even if the posture of the cargo 400 is changed, the processing unit 100 replans the transport route 500 of the unmanned forklift 200. The processing unit 100 constructs a transport route 500 that the unmanned forklift 200 can safely pass through based on a combination of several bounding boxes 301 that can be safely passed. After the screening calculation of the above-mentioned several bounding boxes 301, the processing unit 100 can obtain several bounding boxes 301 that the unmanned forklift 200 can pass through. The processing unit 100 replans the transport route 500 based on the screened several bounding boxes 301. Preferably, the processing unit 100 can store the screened several bounding boxes 301 that enable the unmanned forklift 200 to pass safely, so as to call them during the transportation process of another unmanned forklift 200.

[0086] The present invention can effectively reduce the calculation amount of collision relationship determination and collision check by screening the above-mentioned several boundary boxes 301, and can also serve as a data reserve for the subsequent planning of the transport route 500. During the cargo transportation process of the unmanned forklift 200, it can quickly determine whether the transport route 500 is consistent with the size of the cargo 400, and if it does not match, quickly plan the transport route 500 on which the unmanned forklift 200 can move, thereby improving the efficiency of the unmanned forklift 200 cargo transportation.

[0087] Throughout the text, the features referred to as “preferably” are merely optional and should not be understood as having to be set. Therefore, the applicant reserves the right to abandon or delete the relevant preferred features at any time.

[0088] It should be noted that the above-mentioned specific embodiments are exemplary, and those skilled in the art can come up with various solutions inspired by the disclosure of the present invention, and these solutions also belong to the disclosure scope of the present invention and fall within the protection scope of the present invention. Those skilled in the art should understand that the present invention specification and its drawings are illustrative and do not constitute limitations on the claims. The scope of protection of the present invention is defined by the claims and their equivalents. The present invention specification contains multiple inventive concepts, such as "preferably", "according to a preferred embodiment" or "optionally", all of which indicate that the corresponding paragraph discloses an independent concept, and the applicant reserves the right to file a divisional application based on each inventive concept.

Claims

1. A processing unit for an unmanned forklift, characterized in that: The processing unit (100) is configured to: A minimum cube that can completely contain the goods (400) is used as a bounding box (201), and a virtual channel that can represent the movement trajectory of the unmanned forklift (200) is established in a manner that the bounding box (201) moves along a preset transport route (500); By pre-collision between the virtual channel and the boundary box (301), it is determined whether the unmanned forklift (200) collides with the obstacle (300) under the transport route (500), thereby planning a transport route (500) suitable for the movement of the goods (400) and the unmanned forklift (200).

2. The processing unit according to claim 1, characterized in that The processing unit (100) records each trailing image of the bounding box (201) during its movement in a time-delayed manner, thereby forming a three-dimensional virtual channel.

3. The processing unit according to claim 1 or 2, characterized in that: The processing unit (100) cuts the virtual channel based on the three-dimensional sections of a plurality of bounding boxes (301) to form a plurality of bounding boxes to be measured; The processing unit (100) determines the collision relationship between the bounding box to be tested and the bounding box (301) based on the pre-collision between the bounding box to be tested and the corresponding bounding box (301), so as to screen out a number of bounding boxes to be tested and a number of bounding boxes (301) having a collision relationship.

4. The processing unit according to any one of claims 1 to 3, characterized in that: In the case where the bounding box to be detected has the collision relationship with the bounding box (301), the processing unit (100) divides the irregular bounding box (301) into a plurality of bounding box sub-modules (302) in a segmented and / or split manner, so as to screen out the bounding box sub-modules (302) that may have the collision relationship by pre-colliding the plurality of bounding box sub-modules (302) with the bounding box to be detected.

5. The processing unit according to any one of claims 1 to 4, characterized in that: The processing unit (100) controls the unmanned forklift (200) to change the position of the cargo (400) in a translation or rotation manner based on the three-dimensional collision data, so that the bounding box (201) and the boundary box (301) do not collide.

6. The processing unit according to any one of claims 1 to 5, characterized in that: The processing unit (100) determines, based on the intersection between the bounding box (201) and the boundary box (301) in the three-dimensional collision data, whether the unmanned forklift (200) can avoid colliding with the obstacle (300) after the cargo (400) is translated or rotated.

7. An unmanned forklift, comprising a forklift body and a processing unit (100), characterized in that: The processing unit (100) is configured to: A minimum cube that can completely contain the goods (400) is used as a bounding box (201), and a virtual channel that can represent the movement trajectory of the unmanned forklift (200) is established in a manner that the bounding box (201) moves along a preset transport route (500); By pre-collision between the virtual channel and the boundary box (301), it is determined whether the unmanned forklift (200) collides with the obstacle (300) under the transport route (500), thereby planning a transport route (500) suitable for the movement of the goods (400) and the unmanned forklift (200).

8. The unmanned forklift according to claim 7, characterized in that: The processing unit (100) records each trailing image of the bounding box (201) during its movement in a time-delayed manner, thereby forming a three-dimensional virtual channel.

9. The unmanned forklift according to claim 7 or 8, characterized in that: The processing unit (100) cuts the virtual channel based on the three-dimensional sections of a plurality of bounding boxes (301) to form a plurality of bounding boxes to be measured; The processing unit (100) determines the collision relationship between the bounding box to be tested and the bounding box (301) based on the pre-collision between the bounding box to be tested and the corresponding bounding box (301), so as to screen out a number of bounding boxes to be tested and a number of bounding boxes (301) having a collision relationship.

10. The unmanned forklift according to any one of claims 7 to 9, characterized in that: In the case where the bounding box to be detected has the collision relationship with the bounding box (301), the processing unit (100) divides the irregular bounding box (301) into a plurality of bounding box sub-modules (302) in a segmented and / or split manner, so as to screen out the bounding box sub-modules (302) that may have the collision relationship by pre-colliding the plurality of bounding box sub-modules (302) with the bounding box to be detected.

Images