CN114815851A

CN114815851A - Robot following method, robot following device, electronic device, and storage medium

Info

Publication number: CN114815851A
Application number: CN202210647762.5A
Authority: CN
Inventors: 陈梦婷; 杨淼; 甘泉; 谌振宇
Original assignee: Beijing Haqi Robot Technology Co ltd
Current assignee: Beijing Haqi Robot Technology Co ltd
Priority date: 2022-06-08
Filing date: 2022-06-08
Publication date: 2022-07-29

Abstract

The invention discloses a robot following method, a robot following device, electronic equipment and a storage medium. The method comprises the steps of acquiring motion following information of the robot when the robot follows a target object to move in real time; the motion following information is used for describing the position of a target object under a robot coordinate system, the relative motion relation between the robot and peripheral obstacles, whether the obstacles exist around the robot and the motion state of the robot, and the motion following information is input into a preset speed control model and output to obtain a speed control signal of the robot; the speed control model is obtained by performing reinforcement learning training according to a preset reward function, the robot can be controlled to continue to move along with a target object according to a speed control signal, the robot and the target object are kept within a preset distance range, and the robot is prevented from colliding with surrounding obstacles, so that the problem that the target is easy to miss or lose due to the fact that the following process is easily influenced by interferents in the environment is solved, and the following stability is improved.

Description

Robot following method, robot following device, electronic device, and storage medium

Technical Field

The present invention relates to the field of robot control technologies, and in particular, to a robot following method and apparatus, an electronic device, and a storage medium.

Background

With the continuous emergence of various robots, such as logistics robots, patrol robots, etc., robots are often required to follow to perform tasks.

The robot usually follows in a more complicated environment, and the running environment contains complex factors such as static and dynamic obstacles, targets and non-targets, and the like, but the current scheme mainly adopts laser radar data as input or visual information to navigate and avoid obstacles, and the following process is easily influenced by interferents in the environment, so that the robot has the defects of easily tracking wrong targets or losing targets, and the following stability is reduced.

Disclosure of Invention

The invention provides a robot following method, a robot following device, electronic equipment and a storage medium, which are used for realizing acceleration and deceleration according to a target distance and an angle and keeping a certain distance from a target to finish variable speed following.

According to an aspect of the present invention, there is provided a robot following method, the method including:

determining motion following information when the robot moves along with the target object; the motion following information is used for describing the position of a target object in a robot coordinate system, the relative motion relation between the robot and peripheral obstacles, whether the obstacles exist around the robot or not and the motion state of the robot;

inputting the motion following information into a preset speed control model, and outputting to obtain a speed control signal of the robot; the speed control model is obtained by performing reinforcement learning training according to a preset reward function;

and controlling the robot to continue to move along with the target object according to the speed control signal, so that the robot and the target object are kept within a preset distance range, and the robot is prevented from colliding with surrounding obstacles.

According to another aspect of the present invention, there is provided a robot following device, the device comprising:

the motion following determination module is used for determining motion following information when the robot follows the target object to move; the motion following information is used for describing the position of a target object in a robot coordinate system, the relative motion relation between the robot and peripheral obstacles, whether the obstacles exist around the robot or not and the motion state of the robot;

the speed signal acquisition module is used for inputting the motion following information into a preset speed control model and outputting to obtain a speed control signal of the robot; the speed control model is obtained by performing reinforcement learning training according to a preset reward function;

and the motion following control module is used for controlling the robot to continue to follow the target object to move according to the speed control signal, so that the robot and the target object are kept within a preset distance range and the robot is prevented from colliding with surrounding obstacles.

According to another aspect of the present invention, there is provided an electronic apparatus including:

at least one processor; and

a memory communicatively coupled to the at least one processor; wherein the content of the first and second substances,

the memory stores a computer program executable by the at least one processor, the computer program being executable by the at least one processor to enable the at least one processor to perform the robot following method according to any of the embodiments of the invention.

According to another aspect of the present invention, there is provided a computer-readable storage medium storing computer instructions for causing a processor to implement the robot following method according to any one of the embodiments of the present invention when executed.

According to the technical scheme of the embodiment of the invention, the motion following information of the robot is obtained in real time when the robot follows the target object to move; the motion following information is used for describing the position of a target object under a robot coordinate system, the relative motion relation between the robot and peripheral obstacles, whether the obstacles exist around the robot and the motion state of the robot, and the motion following information is input into a preset speed control model and output to obtain a speed control signal of the robot; the speed control model is obtained by performing reinforcement learning training according to a preset reward function, the robot can be controlled to continue to move along with a target object according to a speed control signal, the robot and the target object are kept within a preset distance range, and the robot is prevented from colliding with surrounding obstacles, so that the problem that the target is easy to miss or lose due to the fact that the following process is easily influenced by interferents in the environment is solved, and the following stability is improved.

It should be understood that the statements in this section do not necessarily identify key or critical features of the embodiments of the present invention, nor do they necessarily limit the scope of the invention. Other features of the present invention will become apparent from the following description.

Drawings

In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings needed to be used in the description of the embodiments will be briefly introduced below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings based on these drawings without creative efforts.

Fig. 1 is a flowchart of a robot following method provided according to an embodiment of the present invention;

FIG. 2 is a schematic diagram of a self-encoder learning process in dynamic vision, which is applicable according to an embodiment of the present invention;

FIG. 3 is a schematic diagram of a process of using a self-encoder in dynamic vision, which is applied according to an embodiment of the present invention;

FIG. 4 is a schematic structural diagram of a robot follower according to an embodiment of the present invention;

fig. 5 is a schematic structural diagram of an electronic device implementing the robot following method according to the embodiment of the present invention.

Detailed Description

In order to make the technical solutions of the present invention better understood, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.

It is noted that the terms "object," "current," "next," and the like in the description and claims of the invention and in the drawings described above are used for distinguishing between similar elements and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used is interchangeable under appropriate circumstances such that the embodiments of the invention described herein are capable of operation in sequences other than those illustrated or described herein. Furthermore, the terms "comprises," "comprising," and "having," and any variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed, but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus.

It will be appreciated that the data involved in the subject technology, including but not limited to the data itself, the acquisition or use of the data, should comply with the requirements of the corresponding laws and regulations and related regulations.

At present, most of path planning takes original image information of a camera as input, so that a machine has functions of learning navigation, obstacle avoidance and the like, and the small part takes laser radar data as input. The map environment causes that the navigation and obstacle avoidance need to build a map in advance, is only suitable for indoor conditions and is not suitable for navigation and obstacle avoidance outdoors; the visual image is directly utilized for modeling, so that the migration from a simulation training environment to a real environment is difficult. The laser radar has high cost and the imaging effect is greatly influenced by weather and illumination conditions. The laser radar is based on an optical detection principle, and laser can penetrate through transparent glass, so that detection omission with certain probability is caused, and the navigation and obstacle avoidance effects of the robot in a real environment are influenced.

The robot following method, apparatus, electronic device, and storage medium provided in the present application are explained in detail below by various embodiments and alternatives thereof.

Fig. 1 is a flowchart of a robot following method, which is applicable to a situation where a robot performs real-time following motion on a target object, and the method may be executed by a robot following device, where the robot following device may be implemented in a form of hardware and/or software, and the robot following device may be configured in any electronic device with a network communication function. As shown in fig. 1, the robot following method in the present embodiment may include the following steps:

and S110, determining motion following information when the robot follows the target object to move.

The motion following information is used for describing the position of the target object in the robot coordinate system, the relative motion relation between the robot and peripheral obstacles, whether the obstacles exist around the robot and the motion state of the robot.

The robot may be a logistics robot, a patrol robot, a household pet robot, etc., and the target object may be a user who has a robot control authority to use the robot or a user who needs the robot to assist in performing a task. When the target object moves (such as walking, running, etc.), the robot is required to follow behind the target object so that the robot can reach the target object as fast as possible when the target object requires the robot to perform a task.

The motion state of the robot can comprise the motion position, the motion speed, the motion attitude, the motion track trend and the like of the robot. The peripheral obstacles may include, but are not limited to, static objects, moving objects, etc. located around the robot.

In an alternative of this embodiment, determining motion following information when the robot follows the motion of the target object may include steps a1-a 2:

and A1, acquiring a front scene depth map of the front driving of the robot, and reconstructing the front scene depth map through a preset encoder to obtain hidden vector characteristics of the front scene.

The robot is provided with the depth camera, and the front scene of the robot can be subjected to image shooting through the depth camera in the moving process of the robot, so that the front scene depth image including the front obstacle of the robot is obtained. A depth image may be referred to as a range image, and refers to an image in which the distance (depth) from a depth camera to each point in a scene is a pixel value.

And A2, determining the motion relation between the robot and the peripheral obstacles according to the hidden vector characteristics of the front scene and the motion state characteristics of the robot.

The preset encoder can be obtained by training through acquiring depth images from a depth camera in a real environment and a simulation environment based on a variational self-encoder or other encoders. The required hidden vector can be obtained by automatically preprocessing the image through the reconstruction function of the variational self-encoder, so that the time consumption of image processing is saved; meanwhile, the simulation and the real depth map are simultaneously modeled to generate the preset variational self-encoder, so that the compatibility of the simulation and the real depth map is realized, and the migration from the simulation to the real depth map is easier.

In an optional manner of this embodiment, reconstructing the depth map of the front scene by using a preset encoder to obtain hidden vector features of the front scene may include steps B1-B3:

and step B1, cutting the front scene depth map into a ground area depth map and a non-ground area depth map.

And step B2, reconstructing the ground region depth map through the encoder corresponding to the ground region to obtain the hidden vector characteristics of the front ground region.

And step B3, reconstructing the depth map of the non-ground area through the encoder corresponding to the non-ground area to obtain hidden vector characteristics of the front non-ground area.

Referring to fig. 2, since the depth map includes a ground area and a non-ground area, the depth distribution of the ground area and the non-ground area has a large difference, and it is difficult to balance the areas having a large depth difference when performing the hidden vector acquisition. Therefore, modeling can be performed respectively for the ground area and the non-ground area, and a variation self-encoder corresponding to the ground area and a variation self-encoder corresponding to the non-ground area are constructed, so that the depth maps of the ground area and the non-ground area can be reconstructed respectively.

Optionally, a training scene depth map (including a real scene and a simulation scene) may be obtained, the training scene depth map may be first divided into a training ground region depth map and a training non-ground region depth map, the training ground region depth map is reconstructed by using a variational self-encoder, the reconstructed implicit vector features are used for decoding to obtain a preprocessed training ground region depth map, and parameters of the variational self-encoder are updated and adjusted by comparing the preprocessed training ground region depth map with the training ground region depth map before preprocessing to obtain a suitable variational self-encoder corresponding to the ground region. Similarly, the same training mode is adopted for the depth map of the non-ground area for training, and the variational self-encoder corresponding to the non-ground area is obtained.

Referring to fig. 3, when reconstructing the front scene depth map, a region segmentation line between the ground region and the non-ground region may be determined, and the front scene depth map is segmented into a ground region depth map and a non-ground region depth map through the region segmentation line. The region dividing line may be predetermined according to the orientation of a depth camera provided on the robot, the height of the depth camera, and an angle between the orientation of the depth camera and the ground. The segmentation line may be a fuzzy boundary determined by identifying depth values of pixels in the depth map of the front scene and performing statistical analysis on the depth values.

In an optional manner of this embodiment, determining the motion relationship between the robot and the peripheral obstacle according to the hidden vector feature of the front scene and the motion state feature of the robot may include the following processes:

inputting the hidden vector characteristics of the ground area, the hidden vector characteristics of the non-ground area and the motion state characteristics of the robot at the current moment into a pre-trained sequence model, and outputting a hidden vector for representing the motion relation between the robot and the peripheral obstacles at the next moment.

And modeling the hidden vectors of the ground area, the hidden vectors of the non-ground area and the motion trail of the machine, which are generated in the depth map reconstruction process, through a sequence model. When the sequence model is used, the sequence model can predict the position of the peripheral obstacle in the future in advance, and a hidden vector for representing the motion relation between the robot and the peripheral obstacle at the next moment is output. The sequence model can be obtained by training with a sequence model structure such as LSTM.

And coding the depth map into an implicit vector describing an external environment by dynamic vision obstacle avoidance in a variable speed following process, and learning the relative motion relation between the robot and the obstacle in the environment from the environmental implicit vector, the motion state and the motion trail of the robot by using a sequence model for follow-up variable power following.

In another alternative of this embodiment, determining motion following information when the robot follows the motion of the target object may include steps C1-C2:

and step C1, determining the distance and the included angle of the target object relative to the robot under the robot coordinate system through a visual positioning mode or an ultra-wideband technology.

And step C2, determining whether an obstacle exists in the preset distance range around the robot through the obstacle detection sensor.

The obstacle detection sensor includes, but is not limited to, an ultrasonic sensor, a laser radar sensor, and the like, and is used for detecting whether an obstacle exists around the robot and the robot needs to avoid the obstacle urgently.

By adopting the mode, a map does not need to be built in advance, and the distance and the angle of the target object relative to the robot body coordinate system can be obtained in real time through a positioning device of visual positioning or ultra wideband technology (UWB), so that the robot can realize acceleration and deceleration according to the distance and the angle of the target, and the robot can keep a certain distance with the target to finish variable speed following. And (3) avoiding obstacles urgently at near places: the method has the advantages that the ultrasonic sensors and other sensors are used for detecting the obstacles in a certain range, the obstacles are urgently avoided when the obstacles appear in a safety range, and the stability is ensured through multi-mode data input.

S120, inputting the motion following information into a preset speed control model and outputting to obtain a speed control signal of the robot; and the preset speed control model is obtained by performing reinforcement learning training according to a preset reward function.

The input for reinforcement learning includes: the method comprises the steps of generating hidden vectors representing the motion relations of a static object, a moving object and a robot in the environment through a sequence model, determining the motion state (including but not limited to linear velocity, angular velocity and the like) of the robot, and determining the position (including distance P and included angle theta) of a target object in a coordinate system of a body of the robot, wherein P represents a distance, theta represents an included angle in the positive direction of an x axis, and sensors such as ultrasonic waves detect whether obstacles exist in a safe distance range.

The output of reinforcement learning includes: control signals to the robot, such as speed commands; when the output control signal is a speed command, the control signal includes a linear speed of the robot in the horizontal axis x and the vertical axis y and an angular speed of the robot in the yaw direction. The speed command issued by the robot may be as follows: the control _ speed _ x is delta _ x _ speed _ x, the control _ speed _ y is delta _ y _ speed _ y, and the control _ w is delta _ w _ speed _ w. Wherein the delta can be adjusted according to the performance and effect of the robot in the real environment.

The preset speed control model is obtained by performing reinforcement learning training according to a preset reward function, and for example, a reinforcement learning algorithm based on an Actor-critic framework can be realized by adopting algorithms such as PPO, TRPO, DDPG, A3C and the like. The preset reward function further comprises a function for indicating that the distance between the robot and the target object is kept within a preset distance range, smoothly changing the speed of the robot in the speed changing process and controlling the included angle between the preset mark direction of the robot and the preset mark direction of the target object to be within the preset range. Reward functions include, but are not limited to, keeping the machine and target at a distance, avoiding collisions, smoothing the motion, making the angle between the target and the fuselage as large as possible, etc.

And S130, controlling the robot to continue to move along with the target object according to the speed control signal, keeping the robot and the target object within a preset distance range, and preventing the robot from colliding with surrounding obstacles.

The robot can keep a certain distance with a moving target through reinforcement learning by a reward function, and automatic smooth acceleration and deceleration movement is realized according to the distance. And in the following process, a proper route is planned according to the fact that the dynamic vision module reacts to the obstacle map in the environment in advance.

According to the technical scheme of the embodiment of the invention, the motion following information of the robot when following the target object is obtained in real time; the motion following information is used for describing the position of a target object under a robot coordinate system, the relative motion relation between the robot and peripheral obstacles, whether the obstacles exist around the robot and the motion state of the robot, and the motion following information is input into a preset speed control model and output to obtain a speed control signal of the robot; the speed control model is obtained by performing reinforcement learning training according to a preset reward function, the robot can be controlled to continue to move along with a target object according to a speed control signal, the robot and the target object are kept within a preset distance range, and the robot is prevented from colliding with surrounding obstacles, so that the problem that the target is easy to miss or lose due to the fact that the following process is easily influenced by interferents in the environment is solved, and the following stability is improved.

Fig. 4 is a block diagram of a robot following device according to an embodiment of the present invention, where the embodiment is applicable to a situation where a robot performs real-time following motion on a target object, the robot following device may be implemented in a form of hardware and/or software, and the robot following device may be configured in any electronic device with a network communication function. As shown in fig. 4, the robot following device in the present embodiment may include the following: a motion follow determination module 410, a velocity signal acquisition module 420, and a motion follow control module 430. Wherein:

a motion following determination module 410, configured to determine motion following information when the robot follows the target object to move; the motion following information is used for describing the position of a target object in a robot coordinate system, the relative motion relation between the robot and peripheral obstacles, whether the obstacles exist around the robot or not and the motion state of the robot;

the speed signal acquisition module 420 is configured to input the motion following information into a preset speed control model, and output a speed control signal of the robot; the speed control model is obtained by performing reinforcement learning training according to a preset reward function;

and a motion following control module 430, configured to control the robot to continue to follow the target object to move according to the speed control signal, so that the robot and the target object are kept within a preset distance range, and the robot is prevented from colliding with surrounding obstacles.

On the basis of the foregoing embodiment, optionally, determining motion following information when the robot follows the target object to move includes:

acquiring a front scene depth map in front of the robot in driving, and reconstructing the front scene depth map through a preset variational self-encoder to obtain front scene hidden vector characteristics;

and determining the motion relation between the robot and the peripheral obstacles according to the hidden vector characteristics of the front scene and the motion state characteristics of the robot.

On the basis of the foregoing embodiment, optionally, reconstructing the depth map of the front scene by using a preset encoder to obtain hidden vector features of the front scene includes:

slicing the front scene depth map into a ground area depth map and a non-ground area depth map;

reconstructing a ground region depth map through a coder corresponding to a ground region to obtain hidden vector characteristics of a front ground region;

and reconstructing the depth map of the non-ground area through an encoder corresponding to the non-ground area to obtain hidden vector characteristics of the front non-ground area.

On the basis of the foregoing embodiment, optionally, determining a motion relationship between the robot and a peripheral obstacle according to the hidden vector feature of the front scene and the motion state feature of the robot includes:

determining the distance and the included angle of a target object relative to the robot under the robot coordinate system through a visual positioning mode or an ultra-wideband technology mode; whether obstacles exist in a preset distance range around the robot or not is determined through the obstacle detection sensor.

On the basis of the foregoing embodiment, optionally, the preset reward function further includes a function for indicating that the distance between the robot and the target object is kept within a preset distance range, smoothly changing the speed of the robot during the speed change process, and controlling an included angle between a preset mark direction of the robot and a preset mark direction of the target object to be within a preset range.

On the basis of the above embodiment, optionally, the obstacle detecting sensor is an ultrasonic sensor and a radar sensor.

The robot following device provided by the embodiment of the invention can execute the robot following method provided by any embodiment of the invention, has corresponding functions and beneficial effects of executing the robot following method, and the detailed process refers to the relevant operation of the robot following method in the embodiment.

FIG. 5 illustrates a schematic diagram of an electronic device 10 that may be used to implement an embodiment of the invention. Electronic devices are intended to represent various forms of digital computers, such as laptops, desktops, workstations, personal digital assistants, servers, blade servers, mainframes, and other appropriate computers. The electronic device may also represent various forms of mobile devices, such as personal digital assistants, cellular phones, smart phones, wearable devices (e.g., helmets, glasses, watches, etc.), and other similar computing devices. The components shown herein, their connections and relationships, and their functions, are meant to be exemplary only, and are not meant to limit implementations of the inventions described and/or claimed herein.

As shown in fig. 5, the electronic device 10 includes at least one processor 11, and a memory communicatively connected to the at least one processor 11, such as a Read Only Memory (ROM)12, a Random Access Memory (RAM)13, and the like, wherein the memory stores a computer program executable by the at least one processor, and the processor 11 can perform various suitable actions and processes according to the computer program stored in the Read Only Memory (ROM)12 or the computer program loaded from a storage unit 18 into the Random Access Memory (RAM) 13. In the RAM 13, various programs and data necessary for the operation of the electronic apparatus 10 can also be stored. The processor 11, the ROM 12, and the RAM 13 are connected to each other via a bus 14. An input/output (I/O) interface 15 is also connected to bus 14.

A number of components in the electronic device 10 are connected to the I/O interface 15, including: an input unit 16 such as a keyboard, a mouse, or the like; an output unit 17 such as various types of displays, speakers, and the like; a storage unit 18 such as a magnetic disk, an optical disk, or the like; and a communication unit 19 such as a network card, modem, wireless communication transceiver, etc. The communication unit 19 allows the electronic device 10 to exchange information/data with other devices via a computer network such as the internet and/or various telecommunication networks.

The processor 11 may be a variety of general and/or special purpose processing components having processing and computing capabilities. Some examples of processor 11 include, but are not limited to, a Central Processing Unit (CPU), a Graphics Processing Unit (GPU), various specialized Artificial Intelligence (AI) computing chips, various processors running machine learning model algorithms, a Digital Signal Processor (DSP), and any suitable processor, controller, microcontroller, or the like. The processor 11 performs the various methods and processes described above, such as the robot following method.

In some embodiments, method XXX may be implemented as a computer program tangibly embodied on a computer-readable storage medium, such as storage unit 18. In some embodiments, part or all of the computer program may be loaded and/or installed onto the electronic device 10 via the ROM 12 and/or the communication unit 19. When the computer program is loaded into the RAM 13 and executed by the processor 11, one or more steps of the robot following method described above may be performed. Alternatively, in other embodiments, the processor 11 may be configured to perform the robot following method by any other suitable means (e.g. by means of firmware).

Various implementations of the systems and techniques described here above may be implemented in digital electronic circuitry, integrated circuitry, Field Programmable Gate Arrays (FPGAs), Application Specific Integrated Circuits (ASICs), Application Specific Standard Products (ASSPs), system on a chip (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and/or combinations thereof. These various embodiments may include: implemented in one or more computer programs that are executable and/or interpretable on a programmable system including at least one programmable processor, which may be special or general purpose, receiving data and instructions from, and transmitting data and instructions to, a storage system, at least one input device, and at least one output device.

A computer program for implementing the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general purpose computer, special purpose computer, or other programmable data processing apparatus, such that the computer programs, when executed by the processor, cause the functions/acts specified in the flowchart and/or block diagram block or blocks to be performed. A computer program can execute entirely on a machine, partly on a machine, as a stand-alone software package partly on a machine and partly on a remote machine or entirely on a remote machine or server.

In the context of the present invention, a computer-readable storage medium may be a tangible medium that can contain, or store a computer program for use by or in connection with an instruction execution system, apparatus, or device. A computer readable storage medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. Alternatively, the computer readable storage medium may be a machine readable signal medium. More specific examples of a machine-readable storage medium would include an electrical connection based on one or more wires, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

To provide for interaction with a user, the systems and techniques described here can be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to a user; and a keyboard and a pointing device (e.g., a mouse or a trackball) by which a user can provide input to the electronic device. Other kinds of devices may also be used to provide for interaction with a user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form, including acoustic, speech, or tactile input.

The systems and techniques described here can be implemented in a computing system that includes a back-end component (e.g., as a data server), or that includes a middleware component (e.g., an application server), or that includes a front-end component (e.g., a user computer having a graphical user interface or a web browser through which a user can interact with an implementation of the systems and techniques described here), or any combination of such back-end, middleware, or front-end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: local Area Networks (LANs), Wide Area Networks (WANs), blockchain networks, and the internet.

The computing system may include clients and servers. A client and server are generally remote from each other and typically interact through a communication network. The relationship of client and server arises by virtue of computer programs running on the respective computers and having a client-server relationship to each other. The server can be a cloud server, also called a cloud computing server or a cloud host, and is a host product in a cloud computing service system, so that the defects of high management difficulty and weak service expansibility in the traditional physical host and VPS service are overcome.

It should be understood that various forms of the flows shown above may be used, with steps reordered, added, or deleted. For example, the steps described in the present invention may be executed in parallel, sequentially, or in different orders, and are not limited herein as long as the desired results of the technical solution of the present invention can be achieved.

The above-described embodiments should not be construed as limiting the scope of the invention. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions may be made in accordance with design requirements and other factors. Any modification, equivalent replacement, and improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A robot following method, characterized in that the method comprises:

2. The method of claim 1, wherein determining motion following information for the robot as it follows the target object comprises:

acquiring a front scene depth map in front of the robot in driving, and reconstructing the front scene depth map through a preset encoder to obtain front scene hidden vector features;

3. The method of claim 2, wherein reconstructing the depth map of the front scene by a preset encoder to obtain hidden vector features of the front scene comprises:

segmenting the front scene depth map into a ground area depth map and a non-ground area depth map;

4. The method of claim 3, wherein determining the motion relationship between the robot and the peripheral obstacle according to the hidden vector feature of the front scene and the motion state feature of the robot comprises:

5. The method of claim 1, wherein determining motion following information for the robot as it follows the target object comprises:

6. The method of claim 1, wherein the predetermined reward function further comprises instructions for indicating that the distance between the robot and the target object is kept within a predetermined distance range, smoothly changing the speed during the speed change of the robot, and controlling an angle between a predetermined sign direction of the robot and a predetermined sign direction of the target object to be within a predetermined range.

7. The method of claim 5, wherein the obstacle detection sensor is an ultrasonic sensor and a radar sensor.

8. A robot following device, characterized in that the device comprises:

9. An electronic device, characterized in that the electronic device comprises:

at least one processor; and

the memory stores a computer program executable by the at least one processor to enable the at least one processor to perform the robot following method of any one of claims 1-7.

10. A computer-readable storage medium storing computer instructions for causing a processor to perform the robot following method of any one of claims 1-7 when executed.