Logistics package sorting and distribution robot and active obstacle avoidance method thereof

Abstract

The invention discloses a logistics package sorting and distribution robot and an active obstacle avoidance method thereof, and relates to the technical field related to robots, the logistics package sorting and distribution robot comprises a wheel type chassis walking mechanism, a laser radar, a first camera, a mechanical arm, a clamping jaw, a second camera and a carriage arranged on the wheel type chassis walking mechanism; the laser radar and the first camera are arranged in front of the wheel type chassis walking mechanism and used for sensing the environment, the laser radar is used for establishing a navigation path, and the first camera is used for recognizing obstacles under the navigation path; one end of the mechanical arm is arranged on the carriage, and the other end is connected with the clamping jaw for grabbing a target package; the second camera is close to the clamping jaw, is arranged on the mechanical arm and is used for detecting target packages under the multiple logistics packages; the robot not only can complete a single specific task, but also can autonomously complete multiple comprehensive tasks without changing any configuration.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of robots, in particular to a logistics parcel sorting and distribution robot and an active obstacle avoidance method thereof. BACKGROUND

[0002] With the rapid development of the express industry, the problem of terminal distribution of logistics has become increasingly prominent. Terminal distribution refers to the distance from the express delivery from the logistics center to the distribution center to the customer. It is the last link in the logistics distribution and the only link directly contacting with customers. At present, a large number of manpower is mainly used to complete this task. If robots are used to replace manual work to complete this task, it will help to reduce the contact between people and reduce the risk of virus transmission. Therefore, it is of great practical significance to study a logistics robot that can replace humans to perform express sorting and distribution tasks.

[0003] Current logistics robots can be broadly divided into three categories: automated guided vehicles, palletizing robots, and sorting robots. According to different application scenarios, automated guided vehicles can be divided into warehouse robots and distribution robots. Automated guided vehicles are high-performance transport robots, mainly used for cargo handling and distribution. Palletizing robots can replace manual work to classify, handle, and load dangerous goods, with much higher efficiency than humans. Sorting robots are robots that use sensors and robotic arms to achieve fast sorting functions, can work 24 hours a day without interruption, and have high sorting accuracy.

[0004] Chinese patent CN114289321B discloses a sorting and distribution integrated logistics robot mechanism, which includes a sorting robot mechanism and a distribution robot mechanism. The sorting robot mechanism includes a sorting robot base, a ring-shaped conveying device, a sorting mechanical arm device with a camera, an RFID scanner device, etc. The distribution robot mechanism includes a chassis, an electromagnetic lock cabinet, a laser radar, and a machine vision device. The present application is a logistics robot mechanism that combines a sorting robot mechanism and a distribution robot mechanism. After the RFID scanner device inside the sorting robot mechanism identifies the express information, it distributes the express to the waiting distribution robot through the sorting mechanical arm inside it, and the distribution robot completes the distribution task. The function of the sorting robot mechanism is equivalent to a small post office. This sorting and distribution integrated logistics robot mechanism can not only realize the function of traditional logistics robot express distribution, but also solve the problem of personnel assistance required by traditional logistics robots.

[0005] The Chinese patent CN113708438B discloses a charging control method, device and equipment of a package delivery robot and a storage medium. The method comprises the following steps: obtaining current state information of the package delivery robot; determining current power of the package delivery robot based on point information; calculating a maximum load corresponding to the package delivery robot by using a preset calculation method based on standard power data and the current power; obtaining historical delivery information corresponding to each package delivery robot, and predicting the number of packages to be delivered in N time periods based on each historical delivery information; determining a target charging time and sequence of the corresponding package delivery robot according to each maximum load and the number of packages to be delivered in the current time period; and charging each package delivery robot according to the target charging time and sequence. The efficient control of the charging of the package delivery robot is realized, and the use rate of the charging pile, the space utilization rate and the delivery efficiency of the package delivery robot are improved.

[0006] The patent CN116224985A discloses a package delivery robot, which belongs to the field of robots. The package delivery robot adopts a split design of upper and lower parts. An upper circuit is arranged in the upper part, and a lower circuit is arranged in the lower part. Then, the upper port and the lower port of the connector are used to assemble the upper part and the lower part to form a complete package delivery robot. In addition, a first laser radar for distance detection in a preset area around the robot is arranged in the upper circuit of the robot, and a second laser radar for distance detection in a blind area around the robot is arranged in the lower circuit. Through the cooperation between the first laser radar and the second laser radar, the robot can realize blind area-free distance detection, so that the robot can conveniently deliver packages from outdoor to the doorsteps of users in a park or a community, or transfer the packages from the users to an intelligent express cabinet outdoors, which has a good market application prospect.

[0007] The current logistics robot mechanism can only cope with one work task, such as sorting task, stacking task and carrying task, and there is no logistics robot mechanism that can cope with multiple tasks at the same time. SUMMARY

[0008] Based on the technical problems existing in the background art, the present application provides a logistics package sorting and delivery robot and an active obstacle avoidance method thereof, which can not only complete a single specific task, but also autonomously complete comprehensive multiple tasks without changing any configuration.

[0009] This invention proposes a logistics parcel sorting and delivery robot, comprising a wheeled chassis walking mechanism, a lidar, a first camera, a robotic arm, a gripper, a second camera, and a carriage mounted on the wheeled chassis walking mechanism; the lidar and the first camera are positioned in front of the wheeled chassis walking mechanism for environmental perception, the lidar is used to establish a navigation path, and the first camera is used to identify obstacles along the navigation path; one end of the robotic arm is mounted on the carriage, and the other end is connected to the gripper for grasping target parcels; the second camera is positioned near the gripper and on the robotic arm for detecting target parcels under multiple logistics parcels.

[0010] Furthermore, the wheeled chassis walking mechanism includes a base plate, drive wheels, and a chassis frame; the lidar and the first camera are both located at the front of the chassis frame and electrically connected to the industrial control computer; the chassis frame is set on the base plate and forms a groove structure with an open top, and a battery, an industrial control computer, an embedded control board, and a driver are arranged inside the groove structure; the industrial control computer and the embedded control board are connected through a communication interface, and the embedded control board is electrically connected to the driver; each driver's power output end is connected to a drive wheel; four mounting wheel grooves are formed through the base plate, and each drive wheel is located in one of the mounting wheel grooves.

[0011] Furthermore, an emergency stop button, a start / stop button, and an ultrasonic sensor are provided on the outer side of the base plate; The emergency stop button and start / stop button are both located at the rear of the chassis frame; The ultrasonic sensors are installed around the chassis frame and electrically connected to the industrial control computer to detect obstacles in the surrounding area.

[0012] Furthermore, the robotic arm consists of multiple arm segments, with a rotational degree of freedom between adjacent segments and a self-rotational degree of freedom within each segment. A six-dimensional force sensor is installed on the segment connected to the gripper, and a second camera is connected to the front end of the six-dimensional force sensor.

[0013] Furthermore, the robotic arm consists of four sections, with the third section having redundant rotational degrees of freedom.

[0014] Furthermore, the gripper is a two-finger gripper, a three-finger gripper, or a five-finger bionic dexterous hand, with multimodal tactile sensors integrated on the fingertips of the gripper.

[0015] Furthermore, a multimodal tactile sensor is disposed on the fingertip surface of the robotic hand. The multimodal tactile sensor includes a flexible substrate, a pressure sensor, an upper electromagnetic insulating film, and a spiral electrode capacitive sensor arranged sequentially from bottom to top. Pressure sensors are used to detect the force applied to an object in contact; Spiral electrode capacitive sensors are used to detect the dielectric properties of objects and combine them with artificial intelligence algorithms to distinguish the material properties of objects.

[0016] Furthermore, the carriage includes a body, a cover, and a cover control system for controlling the opening / closing of the cover; The top of the compartment has an opening, and a cover plate is hinged to the compartment and covers the opening.

[0017] An active obstacle avoidance method for a logistics parcel sorting and delivery robot includes: The robot described above is used; By moving the robot in a dynamic environment to acquire data from LiDAR and the first camera, an initial planned path is established, and obstacles in front are identified and located in real time within a preset range of data information to obtain real-time identification and location results of obstacles in front. Based on the real-time identification and positioning results, it is determined whether the obstacle in front will block the robot from moving according to the initial planned path data; if it will block, the movement path is replanned based on the size of the obstacle in front to avoid collision with the obstacle in front. During the robot's movement, the ultrasonic sensor monitors the distance information of surrounding obstacles in real time. When an obstacle approaches the robot from outside the area of ​​the first camera, the robot will pause. When the obstacle moves away, the robot will continue to move. The area of ​​the first camera is the area of ​​visual information in front of the robot. During the robot's movement, the visual information is analyzed in real time to determine if there are any obstacles. If there are obstacles, the robot enters the obstacle type recognition stage; if there are no obstacles, the robot moves forward according to the original global planning path.

[0018] Furthermore, in obstacle type recognition, neural networks are used to identify obstacle types.

[0019] The advantages of the logistics parcel sorting and delivery robot and its active obstacle avoidance method provided by this invention are: efficient handling is achieved through a wheeled chassis walking mechanism; a robotic arm with at least 6 degrees of freedom enhances the grasping operation range; and a dexterous gripper with multimodal tactile perception enables interactive sensing. Furthermore, this embodiment does not involve complex operating procedures and has strong universality, showing broad application prospects, especially in the fields of intelligent robot operation and human-computer interaction. Attached Figure Description

[0020] Figure 1 This is a schematic diagram of the structure of the present invention; Figure 2 This is a schematic diagram of the structure of the wheeled chassis running mechanism; Figure 3 This is a schematic diagram of the robotic arm. Figure 4This is a structural diagram of the carriage; Figure 5 A schematic diagram of the gripper structure for setting up a multimodal tactile sensor; Among them, 1-wheeled chassis walking mechanism, 2-LiDAR, 3-robotic arm, 4-gripper, 5-second camera, 6-carriage, 7-first camera, 10-ultrasonic sensor, 11-base plate, 12-drive wheel, 13-chassis frame, 14-battery, 15-industrial computer, 16-embedded control board, 17-driver, 18-mounting wheel groove, 19-emergency stop button, 20-start / stop button, 31-first section, 32-second section, 33-third section, 34-fourth section, 35-six-dimensional force sensor, 41-... - Multimodal tactile sensor, 61- Box body, 62- Cover plate, 63- Cover plate control system, 311- Rotational degree of freedom one, 321- Rotational degree of freedom one, 322- Rotational degree of freedom two, 331- Rotational degree of freedom two, 332- Rotational degree of freedom three, 341- Rotational degree of freedom three, 342- Rotational degree of freedom four, 411- Flexible substrate, 412- Pressure sensor, 413- Upper electromagnetic insulating film, 414- Spiral electrode capacitive sensor. Detailed Implementation

[0021] The technical solution of the present invention will now be described in detail through specific embodiments. Many specific details are set forth in the following description to provide a thorough understanding of the invention. However, the present invention can be implemented in many other ways different from those described herein, and those skilled in the art can make similar modifications without departing from the spirit of the invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0022] like Figures 1 to 5 As shown, the present invention proposes a logistics parcel sorting and delivery robot, including a wheeled chassis walking mechanism 1, a lidar 2, a first camera 7, a robotic arm 3, a gripper 4, a second camera 5, and a carriage 6 mounted on the wheeled chassis walking mechanism 1; the lidar 2 and the first camera 7 are located in front of the wheeled chassis walking mechanism 1 for environmental perception, the lidar 2 is used to establish a navigation path, and the first camera 7 is used to identify obstacles under the navigation path; one end of the robotic arm 3 is mounted on the carriage 6, and the other end is connected to the gripper 4 for grasping target parcels; the second camera 5 is located near the gripper 4 and mounted on the robotic arm 3 for detecting target parcels under multiple logistics parcels.

[0023] Using the environmental perception capabilities of LiDAR 2 and the first camera 7, the robot successfully reaches its destination along the navigation path. The second camera 5 detects the target package, and the robotic arm 3 moves the gripper 4 to grasp it. The robot then opens the cargo compartment 6, places the grasped package inside, closes the compartment 6, and initiates its navigation path, planning the delivery route to the destination. It then controls the wheeled chassis walking mechanism 1 to move to the destination. Next, the robot opens the cargo compartment 6 again, retrieves the package from inside, places it at the destination, closes the compartment 6, and returns to its starting point. During the robot's autonomous movement, an active obstacle avoidance method is integrated. This method involves fusing data from multiple sensors, analyzing the obstacles ahead, and then replanning the path to avoid collisions between the robot and obstacles.

[0024] The wheeled chassis walking mechanism 1 of this embodiment achieves efficient material handling; the robotic arm 3 with at least 6 degrees of freedom enhances the grasping operation range; and the dexterous gripper 4 with multimodal tactile sensing enables interactive perception. Furthermore, this embodiment does not involve complex operating procedures and has strong versatility, showing broad application prospects, especially in the fields of intelligent robot operation and human-computer interaction.

[0025] In one embodiment, the wheeled chassis walking mechanism 1 includes a base plate 11, drive wheels 12 and chassis frame 13; The lidar 2 and the first camera 7 are both located in front of the chassis frame 13 and are electrically connected to the industrial control computer 15. The chassis frame 13 is set on the base plate 11 and forms a groove structure with an open top. The groove structure contains a battery 14, an industrial control computer 15, an embedded control board 16, and a driver 17. The industrial control computer 15 is connected to the embedded control board 16 through a communication interface. The embedded control board 16 is electrically connected to the driver 17. Each driver 17 has a drive wheel 12 connected to its power output end. The base plate 11 has four through-holes for mounting wheels 18, and each drive wheel 12 is located in one of the mounting wheel grooves 18.

[0026] First, the base plate 11 has side plates on at least two sides around its perimeter. The base plate 11 is connected to the chassis frame 13 through these side plates. In addition, in order to improve the stability of the chassis frame 13, the middle frame of the chassis frame 13 is connected to the base plate 11 through the column frame, thereby avoiding structural deformation problems in the middle of the chassis frame 13.

[0027] In addition, an emergency stop button 19, a start / stop button 20, and an ultrasonic sensor 10 are provided on the outer side of the base plate 11; the emergency stop button 19 and the start / stop button 20 are both located behind the chassis frame 13; the ultrasonic sensor 10 is located around the chassis frame 13 and is electrically connected to the industrial control computer 15 to detect obstacles around the chassis.

[0028] It is understood that the industrial control computer 15 in this embodiment integrates multi-sensor fusion algorithms, robot motion control algorithms, robotic arm motion grasping control algorithms, and active obstacle avoidance algorithms, etc., to acquire information detected by LiDAR 2, first camera 7, second camera 5, ultrasonic sensor 10, etc., and thus output robotic arm motion grasping commands, automatic obstacle avoidance commands, etc.

[0029] Understandably, the lidar 2, the first camera 7, and the ultrasonic sensor 10 provide data information for the robot's autonomous navigation and movement. The robot's autonomous movement is achieved through the multi-sensor fusion algorithm, robot motion control algorithm, and active obstacle avoidance algorithm of the industrial control computer.

[0030] In one embodiment, the robotic arm 3 is composed of multiple arm segments, with a rotational degree of freedom between adjacent segments and a self-rotational degree of freedom within each segment. A six-dimensional force sensor 35 is provided on the segment connected to the gripper 4, and a second camera 5 is connected to the front end of the six-dimensional force sensor 35.

[0031] In this embodiment, the number of segments of the robotic arm 3 is not limited and can be selected according to actual needs. This embodiment selects a four-segment arm for description: The robotic arm 3 consists of four segments. The connection between the first segment 31 and the second segment 32, the second segment 32 and the third segment 33, and the third segment 33 and the fourth segment 34 each contains a rotational degree of freedom. Each segment has a self-rotational degree of freedom. The self-rotational degree of freedom of the third segment is usually redundant and can be omitted. Therefore, the robotic arm 3 consists of a six-degree-of-freedom or seven-degree-of-freedom robotic arm and a six-dimensional force sensor 35. The front end of the force feedback sensor is connected to a second camera, and a dexterous hand is installed in front of the second camera. That is, a rotational degree of freedom 1 321 is provided between the first section 31 and the second section 32, a rotational degree of freedom 2 331 is provided between the second section 32 and the third section 33, a rotational degree of freedom 341 is provided between the third section 33 and the fourth section 34, a self-rotational degree of freedom 1 311 is provided on the first section 31, a self-rotational degree of freedom 2 322 is provided on the second section 32, a self-rotational degree of freedom 332 is provided on the third section 33, and a self-rotational degree of freedom 4 342 is provided on the fourth section 34.

[0032] After the first section 31, the second section 32, the third section 33, and the fourth section 34 are connected in sequence, the end of the first section 31 is connected to the carriage 6, and the fourth section 34 is connected to the gripper 4. A second camera 5 and a six-dimensional force sensor 35 are installed at the end of the fourth section 34 near the gripper 4, and the second camera 5 is connected to the front end of the six-dimensional force sensor 35.

[0033] It is understood that the gripper 4 in this embodiment is a two-finger gripper, a three-finger gripper, or a five-finger bionic dexterous hand. A multimodal tactile sensor 41 is integrated on the fingertip of the gripper 4. The multimodal tactile sensor 41 is disposed on the surface of the fingertip of the robotic hand. The multimodal tactile sensor 41 includes a flexible substrate 411, a pressure sensor 412, an upper electromagnetic insulating film 413, and a spiral electrode capacitive sensor 414 arranged sequentially from bottom to top. The pressure sensor 412 is used to detect the force applied to the contacting object. The spiral electrode capacitive sensor 414 is used to detect the dielectric properties of the object and, in conjunction with artificial intelligence algorithms, distinguish the material properties of the object.

[0034] The gripper 4 grasps the object, and the multifunctional flexible electronic skin generates an electrical signal upon contact with the object. The electrical signal is transmitted to the industrial control computer through the data acquisition circuit. The industrial control computer 15 can analyze and identify the characteristics of the object through artificial intelligence algorithms, thereby adjusting the adaptive method of the robotic arm to grasp the object.

[0035] In one embodiment, the carriage 6 includes a body 61, a cover 62, and a cover control system 63 for controlling the opening and closing of the cover 62; the body 61 has an opening at the top, and the cover 62 is hinged to the body 61 and covers the opening of the body 61. The vehicle body of this embodiment has a floor area of ​​0.65 meters x 0.58 meters.

[0036] In this embodiment, the multi-sensor used for obstacle avoidance includes a lidar 2, a first camera 7, and an ultrasonic sensor 10. The active obstacle avoidance method collects data information from multiple sensors, establishes an environmental model and plans a path, analyzes obstacle information encountered during movement, and makes decisions based on multi-sensor information fusion using a neural network algorithm to avoid collisions with obstacles. The specific implementation is as follows: S1. The robot moves in a dynamic environment to acquire data information from the LiDAR 2 and the first camera 7, establishes an initial planned path, and performs real-time identification and positioning processing on obstacles in front within a preset range of the data information to obtain real-time identification and positioning results of the obstacles in front; based on the real-time identification and positioning results, it is determined whether the obstacles in front will block the robot from moving according to the initial planned path data; if they will block the robot, the movement path is replanned based on the size of the obstacles in front to avoid collision with the obstacles in front. S2. During the robot's movement, the ultrasonic sensor 10 monitors the distance information of surrounding obstacles in real time. When an obstacle approaches the robot from a region other than the first camera 7, the robot will pause. When the obstacle moves away, the robot continues to move. The first camera 7 region is the visual information region in front of the robot. During the robot's movement, the visual information is analyzed in real time to see if there are obstacles. If there are obstacles, the robot enters the obstacle type recognition stage. If there are no obstacles, the robot moves forward according to the original global planning path.

[0037] Understandably, in obstacle type recognition, neural networks are used to identify obstacle types, obtain obstacle size information, and determine whether it is possible to pass based on the size information. If it is possible to pass, the original path is followed; if it is not possible, the movement path is replanned to achieve active obstacle avoidance.

[0038] As an example: The wheeled chassis walking mechanism 1 occupies an area of ​​0.65m x 0.58m. The lidar 2 and the first camera 7 establish a navigation view. The robotic arm 3 has 6 degrees of freedom and a working radius of 0.6m. The second camera 5 and the gripper 4 accurately grasp the target object. The carriage 6 has a volume of 0.45m x 0.33m x 0.2m and can accommodate multiple delivery objects. The gripper 4 adopts a three-finger bionic dexterous hand.

[0039] Using LiDAR 2 and the first camera 7 for environmental perception, the wheeled chassis walking mechanism 1 moves to the area where the target object is located. The second camera 5 on the robotic arm 3 detects the target object, and the gripper 4 picks up the target object and places it in the compartment 6. Using LiDAR 2 and the first camera 7 for environmental perception again, the wheeled chassis walking mechanism 1 moves to the destination where the target object needs to be transferred. The second camera 5 identifies the target object in the compartment 6, and the gripper 4 picks up the target object and places it at the destination. The above process is repeated to sort and deliver multiple packages.

[0040] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A logistics parcel sorting and delivery robot, characterized in that, It includes a wheeled chassis walking mechanism (1), a lidar (2), a first camera (7), a robotic arm (3), a gripper (4), a second camera (5), and a carriage (6) mounted on the wheeled chassis walking mechanism (1). The lidar (2) and the first camera (7) are located in front of the wheeled chassis walking mechanism (1) for environmental perception. The lidar (2) is used to establish the navigation path, and the first camera (7) is used to identify obstacles under the navigation path. One end of the robotic arm (3) is mounted on the carriage (6), and the other end is connected to the gripper (4) for grabbing the target package; The second camera (5) is located near the gripper (4) and mounted on the robotic arm (3) to detect the target package under multiple logistics packages.

2. The delivery robot according to claim 1, characterized in that, The wheeled chassis walking mechanism (1) includes a base plate (11), drive wheels (12) and chassis frame (13). The lidar (2) and the first camera (7) are both located in front of the chassis frame (13) and are electrically connected to the industrial control computer (15); The chassis frame (13) is set on the base plate (11) and forms a groove structure with an opening at the top. The groove structure contains a battery (14), an industrial computer (15), an embedded control board (16) and a driver (17). The industrial computer (15) is connected to the embedded control board (16) through a communication interface. The embedded control board (16) is electrically connected to the driver (17). The power output end of each driver (17) is connected to the drive wheel (12). Four mounting wheel grooves (18) are provided through the base plate (11). Each drive wheel (12) is located in the mounting wheel groove (18).

3. The delivery robot according to claim 2, characterized in that, An emergency stop button (19), a start / stop button (20), and an ultrasonic sensor (10) are provided on the outside of the base plate (11). The emergency stop button (19) and start / stop button (20) are both located at the rear of the chassis frame (13); The ultrasonic sensor (10) is located around the chassis frame (13) and electrically connected to the industrial control computer (15) to detect obstacles around the chassis.

4. The delivery robot according to claim 1, characterized in that, The robotic arm (3) consists of multiple arm segments, with one rotational degree of freedom between adjacent segments and one self-rotational degree of freedom within each segment. A six-dimensional force sensor (35) is provided on the segment connected to the gripper (4), and a second camera (5) is connected to the front end of the six-dimensional force sensor (35).

5. The delivery robot according to claim 4, characterized in that, The robotic arm (3) consists of four sections, with the third section having redundant degrees of freedom for rotation.

6. The delivery robot according to claim 1, characterized in that, The gripper (4) is a two-finger gripper, a three-finger gripper, or a five-finger bionic dexterous hand, and a multimodal tactile sensor (41) is integrated on the fingertip of the gripper (4).

7. The delivery robot according to claim 6, characterized in that, A multimodal tactile sensor (41) is disposed on the fingertip surface of the robotic hand. The multimodal tactile sensor (41) includes a flexible substrate (411), a pressure sensor (412), an upper electromagnetic insulating film (413), and a spiral electrode capacitive sensor (414) arranged sequentially from bottom to top. The pressure sensor (412) is used to detect the force applied to a contacting object; The spiral electrode capacitive sensor (414) is used to detect the dielectric properties of an object and to distinguish the material properties of the object by combining artificial intelligence algorithms.

8. The delivery robot according to claim 1, characterized in that, The carriage (6) includes a body (61), a cover (62), and a cover control system (63) for controlling the opening and closing of the cover (62). The top of the compartment (61) has an opening, and a cover plate (62) is hinged to the compartment (61) and covers the opening of the compartment (61).

9. An active obstacle avoidance method for a logistics parcel sorting and delivery robot, characterized in that, include: The robot described in any one of claims 1 to 8 is employed; By moving the robot in a dynamic environment to obtain data information from the lidar (2) and the first camera (7), an initial planned path is established, and real-time identification and positioning processing is performed on obstacles in front within the preset range of the data information to obtain the real-time identification and positioning results of the obstacles in front. Based on the real-time identification and positioning results, it is determined whether the obstacle in front will block the robot from moving according to the initial planned path data; if it will block, the movement path is replanned based on the size of the obstacle in front to avoid collision with the obstacle in front. During the robot's movement, the ultrasonic sensor (10) monitors the distance information of surrounding obstacles in real time. When an obstacle approaches the robot from a region other than the first camera (7), the robot will pause. When the obstacle moves away, the robot continues to move. The region of the first camera (7) is the visual information region in front of the robot. During the robot's movement, the visual information is analyzed in real time to determine if there are any obstacles. If there are obstacles, the robot enters the obstacle type recognition stage; if there are no obstacles, the robot moves forward according to the original global planning path.

10. The active obstacle avoidance method according to claim 9, characterized in that, In obstacle type recognition, neural networks are used to identify obstacle types.

Images