Unmanned forklift based on machine vision

Abstract

The utility model relates to an unmanned forklift based on machine vision. The unmanned forklift comprises a chassis, a multi-degree-of-freedom pallet fork, a cargo identification and detection module, a control system and a power supply module. The unmanned forklift can automatically complete carrying, loading and unloading tasks, so that the labor cost can be reduced, and the production benefit can be improved; through automatic control and intelligent management, manual operation errors can be reduced, so that the accident risk is reduced, and the operation safety is improved.

Description

Technical Field

[0001] The utility model relates to the technical field of forklifts, in particular to an unmanned forklift based on machine vision. Background Art

[0002] Nowadays, with the rapid development of artificial intelligence, sensors, computers and other fields, unmanned driving has made breakthrough progress in all aspects, and the automation level of the logistics industry is also continuously improving. As one of the important tools in logistics operations, forklifts have the characteristics of large load capacity, high degree of standardization, and wide application in various industries. Therefore, their intelligent transformation plays a vital role in realizing the comprehensive automation of logistics in future factories in my country. Forklifts equipped with unmanned driving technology can autonomously perform tasks such as handling and loading and unloading of goods in environments such as factories, warehouses or logistics centers. Smart forklifts are usually equipped with a variety of sensors that can identify obstacles in the environment, follow predetermined paths or autonomously plan paths, so as to complete tasks autonomously. This automation system can effectively improve logistics efficiency, reduce labor costs, and help the innovative development of the manufacturing industry. Therefore, it is widely used in modern logistics and manufacturing industries.

[0003] Faced with the opportunities brought by the development of a new generation of artificial intelligence and the increasing demand for intelligent loading and unloading of containers, there is an urgent need for an unmanned forklift that can be used in container loading and unloading scenarios. Summary of the invention

[0004] The purpose of the utility model is to overcome the deficiencies of the prior art and provide an unmanned forklift based on machine vision.

[0005] In order to achieve the above purpose, the utility model adopts the following technical solutions:

[0006] An unmanned forklift based on machine vision, including a chassis, a multi-degree-of-freedom fork, a cargo identification and detection module, a control system, and a power supply module;

[0007] The multi-degree-of-freedom fork has three degrees of freedom: lifting, left and right lateral movement, and pitching. The multi-degree-of-freedom fork is connected to the front end of the chassis, and the multi-degree-of-freedom fork comprises a fork, a fork base, a fork tilting seat, a lifting slide rail truss structure and a transverse slide rail truss structure; the fork base is fixed to the front end of the chassis, and the two sides of the fork tilting seat are respectively fixed with inclination shafts, and the inclination shafts on both sides are respectively rotatably sleeved with the base bearing seats corresponding to the two sides of the fork base, and a inclination motor is fixed on one end of the fork base, and the inclination motor is transmission-connected with the inclination shaft on one side, and the inclination shaft on the other side is connected with an inclination encoder for obtaining the inclination state of the fork tilting seat; the lifting slide rail truss structure is arranged on the top of the fork tilting seat, and the transverse slide rail truss structure is connected to the lifting slide rail truss structure, and the lifting slide rail truss structure drives the transverse slide rail truss structure to perform lifting and lowering movements; the fork is connected to the transverse slide rail truss structure, and the transverse slide rail truss structure drives the fork to move left and right;

[0008] The cargo identification and detection module includes a laser radar and a visual recognition camera;

[0009] The control system and the power supply module are respectively located above the chassis. The control system is used to control the entire forklift to work automatically, and the power supply module is used to supply power to the entire forklift.

[0010] Furthermore, four top corners of the bottom plate of the chassis are respectively connected to McRae wheels, and the driving shaft of each McRae wheel is driven to rotate by a corresponding driving motor through a synchronous belt structure.

[0011] Furthermore, the driving shaft of the McRae wheel is assembled on a vertical bearing seat, and the bearing seat is installed on the aluminum profile of the chassis close to the positioning edge of the McRae wheel positioning base plate; the shaft ends of each McRae wheel driving shaft are respectively connected to a wheel axle encoder to obtain the movement status of the four wheels in real time.

[0012] Furthermore, ultrasonic distance measuring sensors are arranged on both sides and the rear end of the chassis to sense obstacles around the forklift.

[0013] Furthermore, the lifting slide rail truss structure includes an inner truss and an outer truss, and the two sides of the transverse truss fixing seat connected to the transverse slide rail truss structure are respectively slidably fitted on the inner truss through sliders; the outer truss is fixed on the top of the fork tilting seat, and the inner truss and the outer truss are also slidably connected by sliders; the lifting motor is installed on the transverse truss fixing seat, and the two ends of the first-level lifting synchronous belt are respectively connected to the inner truss connecting plates at the upper and lower ends of the inner truss, and the middle part of the first-level lifting synchronous belt is converted through the idle wheel and then wound around the synchronous wheel at the front end of the lifting motor; both ends of the second-level lifting synchronous belt are connected to the top of the outer truss, and the middle part of the second-level lifting synchronous belt is also connected to the transverse truss fixing seat; a lifting encoder is provided above the outer truss to obtain the lifting displacement information of the fork.

[0014] Furthermore, the transverse slide rail truss structure includes a transverse truss, in which the fork is slidably mounted through a slider, a transverse motor is fixed at one end of the transverse truss, and a transverse encoder is arranged at the other end to feedback the lateral displacement information of the fork; the transverse synchronous belt is fixed on the fork, and the transverse motor drives the lateral movement of the fork through the synchronous belt.

[0015] Furthermore, protective baffles are arranged around the chassis.

[0016] Furthermore, the power supply module is a 24V DC power supply module.

[0017] The utility model adopts the above technical scheme, and has the following beneficial effects: by automatically identifying and locating containers and goods, and autonomously planning the optimal path, the loading and unloading efficiency can be greatly improved, thereby shortening the logistics time and improving the logistics efficiency; because the unmanned forklift can automatically complete the handling and loading and unloading tasks, it can reduce labor costs and improve production efficiency; through automated control and intelligent management, human operational errors can be reduced, thereby reducing the risk of accidents and improving operational safety. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] The present invention is further described in detail below with reference to the accompanying drawings and specific implementations:

[0019] Figure 1 It is a schematic diagram of the overall structure of the utility model;

[0020] Figure 2 It is a schematic diagram of the connection structure between the fork tilting seat and the fork base;

[0021] Figure 3 is a schematic diagram of a lifting slide rail truss structure;

[0022] Figure 4 is a schematic diagram of a lifting slide rail truss structure;

[0023] Figure 5 This is the flow chart of the fusion algorithm. DETAILED DESCRIPTION

[0024] like Figure 1-4 As shown, the unmanned forklift based on machine vision of the utility model includes a chassis 1, a multi-degree-of-freedom fork 2, a cargo identification and detection module, a control system 3 and a power supply module 4.

[0025] In order to ensure that the chassis 1 has sufficient flexibility, a four-McRam wheel omnidirectional steering mechanism is adopted. The main frame of the chassis 1 is composed of an aluminum profile frame and a bottom plate; the four top corners of the bottom plate of the chassis 1 are respectively connected to the McRam wheels 7, and the drive shaft of each McRam wheel 7 is driven to rotate by the corresponding drive motor through the synchronous belt structure. The drive shaft of the McRam wheel 7 is assembled on the vertical bearing seat, and the vertical bearing seat is installed on the aluminum profile of the chassis 1 close to the positioning edge of the positioning bottom plate of the McRam wheel 7 to reduce the installation deviation of the four wheels and ensure the accuracy of the omnidirectional movement positioning of the forklift. The shaft ends of the drive shafts of each McRam wheel 7 are respectively connected to the wheel axle encoder to obtain the movement status of the four wheels in real time. The chassis 1 is arranged with protective guards 8 around it, and the two sides and rear end of the chassis 1 (protective guard 8) are arranged with ultrasonic ranging sensors 9 to sense obstacles around the forklift to avoid damage caused by collision between the model and the obstacles. A control system 3 is installed above the chassis 1. The control system 3 is used to control the automatic operation of the entire forklift. The control system 3 is located inside a control box. The electronic circuit boards of the entire vehicle are installed in this box. All line signals are also concentrated in the box for processing. The box is equivalent to the "brain" of the forklift. The entire forklift is powered by a 24V DC power supply module. The power supply module 4 is installed at the upper tail of the chassis 1, which can also serve as a counterweight for the forklift.

[0026] Since the forklift needs to accurately identify and locate the fork hole corresponding to the goods when transporting goods, it is difficult to locate only by the omnidirectional movement of the McRae wheel 7, and it is easily limited by space. Therefore, the fork 21 at the front end of the chassis 1 is modified so that the fork 21 has three degrees of freedom: lifting, left and right lateral movement, and pitching. After the forklift identifies and moves to the side of the goods for rough positioning, the position of the fork point can be transmitted to the control system 3 by relying on the visual system, and precise positioning can be performed by relying on the movement of the fork 21.

[0027] Specifically, the multi-degree-of-freedom fork 2 is connected to the front end of the chassis 1, and the multi-degree-of-freedom fork 2 includes a fork 21, a fork base 22, a fork tilting seat 23, a lifting slide rail truss structure and a transverse slide rail truss structure; the fork base 22 is fixed to the front end of the chassis 1, and the two sides of the fork tilting seat 23 are respectively fixed with inclination shafts 231, and the inclination shafts 231 on both sides are respectively rotatably sleeved with the corresponding base bearing seats on both sides of the fork base 22, so that the fork tilting seat 23 can rotate relative to the fork base 22. Most of the overall parts of the fork 21 are fixed on the fork tilting seat 23, so the inclination adjustment function of the fork 21 is realized as the fork tilting seat 23 rotates. A tilt motor 232 is fixed on one end of the fork base 22, and the tilt motor 232 is transmission-connected to the tilt shaft 231 on one side, and the tilt shaft 231 on the other side is connected to a tilt encoder 233 for obtaining the tilt state of the fork tilt seat 23, and a small micro switch is used as a limit point signal.

[0028] The lifting slide rail truss structure is arranged on the top of the fork tilting seat 23, the transverse slide rail truss structure is connected to the lifting slide rail truss structure, and the lifting slide rail truss structure drives the transverse slide rail truss structure to perform lifting and lowering movements; the fork 21 is connected to the transverse slide rail truss structure, and the transverse slide rail truss structure drives the fork 21 to move horizontally left and right.

[0029] In order to better stack the goods, the fork 21 is designed to be able to be lifted and lowered. The lifting adopts a synchronous belt transmission mechanism with a double stroke. Specifically, the lifting slide rail truss structure includes an inner truss 24 and an outer truss 25. The lifting inner truss 24 and outer truss 25 are in a two-stage telescopic manner. The two sides of the transverse truss fixing seat 26 connected to the transverse slide rail truss structure are respectively slidably fitted on the inner truss 24 through sliders, and the transverse truss fixing seat 26 can move inside the inner truss 24. This is the first stage of lifting; the outer truss 25 is fixed on the top of the fork tilting seat 23, and the inner truss 24 and the outer truss 25 are also slidably connected by sliders, so that the inner truss 24 can slide up and down inside the outer truss 25, which is the second stage of lifting mechanism. The lifting motor 27 is installed on the transverse truss fixing seat 26, and the two ends of the first-level lifting synchronous belt 28 are respectively connected to the inner truss connecting plates 211 at the upper and lower ends of the inner truss 24. The middle part of the first-level lifting synchronous belt 28 is wound around the synchronous wheel at the front end of the lifting motor 27 after the direction is changed by the idler wheel; when the lifting motor 27 rotates, since the two ends of the first-level lifting synchronous belt 28 are fixed, the transverse truss fixing seat 26 will slide in the inner truss 24, and the fork 21 will be driven to achieve the first level of lifting. Both ends of the secondary lifting synchronous belt 29 are connected to the top of the outer truss 25, and the middle part of the secondary lifting synchronous belt 29 is also connected to the transverse truss fixing seat 26; therefore, as the transverse truss fixing seat 26 moves, the inner truss 24 will move relative to the outer truss 25, thereby achieving the second level of lifting. A lifting encoder 210 is provided above the outer truss 25 to obtain the lifting displacement information of the fork 21.

[0030] The transverse slide rail truss structure includes a transverse truss 212, and the fork 21 is slidably mounted in the transverse truss 212 through a slider. A transverse motor 213 is fixed to one end of the transverse truss 212, and a transverse encoder 214 is arranged at the other end to feedback the lateral displacement information of the fork 21; a transverse synchronous belt 215 is fixed on the fork 21, and the transverse motor 213 drives the lateral movement of the fork 21 through the synchronous belt.

[0031] The cargo identification and detection module includes a laser radar 5 and a visual recognition camera 6. The module adopts a relatively mature hardware architecture to achieve synchronous calibration and fusion of multiple sensors, and can use the fused data to detect and identify target cargo. The point cloud data obtained by the laser radar 5 is used to build the map scene of the unmanned forklift operation; at the same time, the model also uses advanced path planning and obstacle avoidance algorithms, which can be applied to specific scenarios and needs.

[0032] The detection and identification of goods requires the collection of corresponding data. The selection and preparation of data sets are very important for the success of cargo detection and identification. In order to achieve good recognition results under different conditions, the data set should be diverse and representative, including target objects of different types, postures, lighting conditions and backgrounds. The unmanned forklift mainly uses YOLO as the main target detection algorithm, which converts the target detection task into a regression problem. Since the prediction of the YOLO algorithm is based on the entire image, all detected target signals, including their categories and locations, can be output at one time, which meets the operating conditions of unmanned forklifts in container scenarios.

[0033] The specific process of unmanned forklift operation path planning is as follows: First, a model representing the forklift's working environment needs to be established. This model includes all possible paths, obstacles, target locations and other information. For each target location, that is, the location of different goods, use a suitable algorithm to find a feasible path from the starting point to the target location; for each searched path, evaluate its feasibility, efficiency and other factors, determine its priority, and select an optimal path as the forklift's movement plan based on the evaluation results. If a single path algorithm is used, there will be certain limitations. At present, it is necessary to collect information through the fusion of multiple types of sensors to implement algorithm optimization. Considering that the unmanned forklift model needs to avoid obstacles and moving objects, it is planned to use the A* fusion DWA algorithm. The process of the proposed optimization fusion algorithm is as follows Figure 5 As shown in the figure. Since the A* fusion DWA algorithm is simple, easy to understand and implement, and has a fast calculation speed, it is suitable for small-scale grid path planning and is more suitable for the path planning of unmanned forklifts in container loading and unloading scenarios. The optimized path algorithm enables the unmanned forklift model to search for suitable target goods in the optimal path according to the next stacking space, thereby improving the handling efficiency of the forklift.

[0034] In summary, the key technologies for the development of unmanned forklifts based on machine vision include the construction of an unmanned forklift experimental platform; the underlying fusion of visual recognition cameras and lidar sensors, the collection and recognition of signals; programming to complete target detection and path planning; and the implementation of an unmanned forklift system based on Python. After the single-chip microcomputer collects information from the sensor, it controls the movement of the intelligent forklift, the lifting, tilting and lateral movement of the fork through the motor, and completes the transportation and loading and unloading of goods by combining target detection and path planning. The key point of the forklift to achieve unmanned operation process technology lies in the identification and fork positioning of the target goods and the autonomous planning of the intelligent forklift operation path. The research value of unmanned forklift operation lies in the automatic identification and positioning of containers and goods, as well as the autonomous planning of the optimal path, which can greatly improve the loading and unloading efficiency, thereby shortening the logistics time and improving the logistics efficiency; since the unmanned forklift can automatically complete the handling and loading and unloading tasks, it can reduce labor costs and improve production efficiency; through automated control and intelligent management, human operation errors can be reduced, thereby reducing the risk of accidents and improving operation safety.

[0035] The above describes the specific implementation of the present invention, but those skilled in the art should understand that this is only an example. Those skilled in the art can make various changes or modifications to this implementation without departing from the principle and essence of the present invention, but these changes and modifications are all within the scope of protection of the present invention.

Claims

1. Unmanned forklift based on machine vision, characterized by: It includes chassis, multi-degree-of-freedom forks, cargo identification and detection modules, control systems and power supply modules; The multi-degree-of-freedom fork has three degrees of freedom: lifting, left and right lateral movement, and pitching. The multi-degree-of-freedom fork is connected to the front end of the chassis, and the multi-degree-of-freedom fork comprises a fork, a fork base, a fork tilting seat, a lifting slide rail truss structure and a transverse slide rail truss structure; the fork base is fixed to the front end of the chassis, and the two sides of the fork tilting seat are respectively fixed with inclination shafts, and the inclination shafts on both sides are respectively rotatably sleeved with the base bearing seats corresponding to the two sides of the fork base, and a inclination motor is fixed on one end of the fork base, and the inclination motor is transmission-connected with the inclination shaft on one side, and the inclination shaft on the other side is connected with an inclination encoder for obtaining the inclination state of the fork tilting seat; the lifting slide rail truss structure is arranged on the top of the fork tilting seat, and the transverse slide rail truss structure is connected to the lifting slide rail truss structure, and the lifting slide rail truss structure drives the transverse slide rail truss structure to perform lifting and lowering movements; the fork is connected to the transverse slide rail truss structure, and the transverse slide rail truss structure drives the fork to move left and right; The cargo identification and detection module includes a laser radar and a visual recognition camera; The control system and the power supply module are respectively located above the chassis. The control system is used to control the entire forklift to work automatically, and the power supply module is used to supply power to the entire forklift.

2. The unmanned forklift based on machine vision according to claim 1, characterized in that: The four top corners of the bottom plate of the chassis are respectively connected with McRae wheels, and the driving shaft of each McRae wheel is driven to rotate by the corresponding driving motor through a synchronous belt structure.

3. The unmanned forklift based on machine vision according to claim 2, characterized in that: The driving shaft of the McRae wheel is assembled on a vertical bearing seat, and the bearing seat is installed on the aluminum profile of the chassis close to the positioning edge of the McRae wheel positioning base plate; the shaft end of each McRae wheel driving shaft is respectively connected to a wheel axle encoder to obtain the movement status of the four wheels in real time.

4. The unmanned forklift based on machine vision according to claim 1, characterized in that: Ultrasonic distance measuring sensors are arranged on both sides and the rear end of the chassis to sense obstacles around the forklift.

5. The unmanned forklift based on machine vision according to claim 1, characterized in that: The lifting slide rail truss structure includes an inner truss and an outer truss, and the two sides of the transverse truss fixing seat connected to the transverse slide rail truss structure are respectively slidably fitted on the inner truss through sliders; the outer truss is fixed on the top of the fork tilting seat, and the inner truss and the outer truss are also slidably connected by sliders; the lifting motor is installed on the transverse truss fixing seat, and the two ends of the first-level lifting synchronous belt are respectively connected to the inner truss connecting plates at the upper and lower ends of the inner truss, and the middle part of the first-level lifting synchronous belt is converted through the idle wheel and then wound around the synchronous wheel at the front end of the lifting motor; both ends of the second-level lifting synchronous belt are connected to the top of the outer truss, and the middle part of the second-level lifting synchronous belt is also connected to the transverse truss fixing seat; a lifting encoder is provided above the outer truss to obtain the lifting displacement information of the fork.

6. The unmanned forklift based on machine vision according to claim 1, characterized in that: The transverse slide rail truss structure includes a transverse truss, in which the fork is slidably mounted through a slider, a transverse motor is fixed at one end of the transverse truss, and a transverse encoder is arranged at the other end to feedback the lateral displacement information of the fork; a transverse synchronous belt is fixed on the fork, and the transverse motor drives the lateral movement of the fork through the synchronous belt.

7. The unmanned forklift based on machine vision according to claim 1, characterized in that: Protective baffles are arranged around the chassis.

8. The unmanned forklift based on machine vision according to claim 1, characterized in that: The power supply module is a 24V DC power supply module.

Images