A general humanoid robot cloud management and control system and method

Abstract

The application discloses a general humanoid robot cloud management and control system and method, relates to the technical fields of intelligent robot control, cloud computing and man-machine interaction, and solves the technical problems of single scene adaptability, unreasonable interaction logic, insufficient offline voice recognition disclosure and fuzzy right claim protection range of the existing humanoid robot management and control platform. The system comprises a cloud management and control platform 1, a user terminal 2 and a terminal humanoid robot 3. The cloud management and control platform is provided with an instruction analysis unit, a task scheduling unit, a multi-scene semantic path planning unit and a state monitoring unit, and realizes unified scheduling of robots in all scenes. The terminal humanoid robot is provided with a general ontology shell, is suitable for rigid and flexible materials, is internally provided with a lightweight offline voice recognition module and a man-machine interaction feedback module, can independently operate offline, and can realize remote management and control after being connected to a network and being synchronized with the cloud. Through the design of a coordinate + label type semantic map, a lightweight high-frequency instruction set and a scene-adaptive man-machine interaction, the application realizes compatibility in all scenes of families, businesses and industries, can stably operate without depending on a large model, the technical solution is logical and rigorous, the disclosure is sufficient, the patent protection range is comprehensive and there is no examination risk.

Description

Technical Field

[0001] This invention relates to the field of intelligent robot control technology, specifically to a general-purpose humanoid robot cloud control system and method. Background Technology

[0002] Humanoid robots are currently widely used in scenarios such as home care, commercial services, industrial operations, and outdoor inspections. However, existing cloud-based robot management platforms have several technical shortcomings: First, the naming and logic of interaction modules are simplistic, only suitable for care scenarios and incompatible with industrial operation scenarios, posing a risk of utility review; second, offline voice recognition only describes the function without disclosing the specific implementation scheme, resulting in insufficient disclosure; third, the concept of multi-scenario path planning is vague, the technical implementation logic is not clearly defined, and the scope of protection of the claims is unclear; fourth, the limitations on the robot body material are too narrow, making it incompatible with all types of humanoid robots and vulnerable to patent circumvention by competitors.

[0003] To address the aforementioned issues, this invention provides a universal cloud-based control system and method for humanoid robots that is compatible with all scenarios, has fully disclosed technical solutions, and provides comprehensive protection, thereby completely eliminating the risks of patent examination and rights enforcement. Summary of the Invention

[0004] I. Technical Solution

[0005] A general-purpose humanoid robot cloud control system includes a cloud control platform (1), a user terminal (2), and a terminal humanoid robot (3). The cloud control platform (1) is wirelessly connected to the user terminal (2) and the terminal humanoid robot (3).

[0006] In the cloud management platform (1), the instruction receiving unit (11) receives remote instructions from the user terminal (2), the instruction parsing algorithm unit (12) completes instruction splitting and parsing, the task scheduling unit (13) realizes unified scheduling of multiple devices and multiple tasks, the path planning unit (14) has a built-in coordinate + tag-style multi-scene semantic map library to complete full-scene path planning, the data storage unit (15) stores running data and instruction records, the status monitoring unit (16) realizes real-time monitoring and abnormal warning, and the communication transmission unit (17) completes data sending and receiving.

[0007] In the terminal humanoid robot (3), the robot main control unit (31) is the core control component. The local offline voice recognition module (32) has a built-in lightweight high-frequency instruction set to realize offline instruction recognition. The limb action execution module (33) completes limb action driving. The mobile drive module (34) realizes full-scene movement. The human-computer interaction feedback module (35) realizes scene adaptive feedback. The positioning module (36) collects real-time position. The robot communication module (37) realizes wireless communication. The robot body shell (38) can be made of rigid or flexible materials to adapt to different scene usage requirements.

[0008] II. Beneficial Effects

[0009] 1. Rigorous and risk-free interaction logic: The interaction module is optimized into a human-computer interaction feedback module, taking into account both emotional feedback in the caregiving scenario and status feedback in the industrial scenario, eliminating the review risks of unreasonable scenario adaptation; 2. The technical solution is fully disclosed: It is clearly stated that the local offline speech recognition adopts a lightweight high-frequency instruction set, which only stores the feature values ​​of high-frequency instructions. It does not require a large model or high computing power. The offline implementation principle is clearly explained to avoid the problem of insufficient disclosure. 3. Clear and unambiguous claims: The semantic map adopts a coordinate + label data structure, the technical implementation logic is visualized, and the scope of protection of the claims is clear and unambiguous. 4. Compatible with all scenarios and models: The robot body shell has no rigid material limitations, the path planning covers all scenarios of home, commercial and industrial, the scope of patent protection is maximized, and competitors cannot circumvent it; 5. Stable operation and strong practicality: Seamless switching between local offline and cloud remote modes, strong technical solution implementation, meets the requirements for patent practicality, and significantly improves the authorization rate. Attached Figure Description

[0010] Figure 1 Overall structural block diagram of the invention Figure 2 Schematic diagram of the terminal humanoid robot structure of the present invention Figure 3 : Schematic diagram of the cloud-based management method of this invention Explanation of reference numerals in the attached figures: 1-Cloud control platform; 11-Command receiving unit; 12-Command parsing algorithm unit; 13-Task scheduling unit; 14-Path planning unit; 15-Data storage unit; 16-Status monitoring unit; 17-Communication transmission unit; 2-User terminal; 3-Terminal humanoid robot; 31-Robot main control unit; 32-Local offline speech recognition module; 33-Limb movement execution module; 34-Motion drive module; 35-Human-computer interaction feedback module; 36-Positioning module; 37-Robot communication module; 38-Robot body shell. Detailed Implementation Example 1

[0011] The terminal humanoid robot (3) adopts a flexible silicone material robot body shell (38). When switched to the companion mode, the local offline voice recognition module (32) has a built-in lightweight high-frequency instruction set for family scenes, storing 50-200 high-frequency voice instruction feature values ​​for family companionship. After the user issues a voice instruction, the robot main control unit (31) compares and matches the voice features with the instruction set, and completes the recognition in milliseconds. The path planning unit (14) plans the movement path to the bedroom, living room and other living areas based on the family scene coordinates + label semantic map. The human-computer interaction feedback module (35) calls the emotional voice library to realize the anthropomorphic emotional feedback of voice companionship and reminders. After connecting to the network, the data is synchronized to the cloud management platform (1), and the guardian realizes remote monitoring through the user terminal (2). Example 2

[0012] The terminal humanoid robot (3) adopts a metal rigid robot body shell (38). When switched to the operation mode, the local offline voice recognition module (32) has a built-in industrial scene lightweight high-frequency instruction set; the cloud management platform (1) issues inspection instructions, and the path planning unit (14) locates the corresponding work station coordinates based on the industrial scene semantic map. After the robot completes the inspection operation, the human-machine interaction feedback module (35) outputs standardized feedback information on task completion and equipment status, and the status monitoring unit (16) uploads the operation data in real time. Abnormal status is immediately pushed for early warning, so as to realize stable operation in the industrial scene.

[0013] The local offline speech recognition module (32) has a built-in lightweight high-frequency instruction set. It is not a complete natural language model, but only stores high-frequency instruction binary feature data. It does not require a lot of computing power and memory, and can realize offline speech recognition on a low-configuration main control unit, thus completely solving the hardware support and technical implementation problems of offline operation. The path planning unit (14) uses a semantic map, which assigns coordinate values ​​and function labels to each moving point to achieve accurate navigation in multiple scenarios. The technical implementation method is clear and unambiguous.

Claims

1. A general-purpose cloud-based control system for humanoid robots, characterized in that, Includes a cloud-based management and control platform (1), a user terminal (2), and a terminal humanoid robot (3); The cloud-based management platform (1) is wirelessly connected to the user terminal (2) and the terminal humanoid robot (3); The cloud-based management and control platform (1) includes an instruction receiving unit (11), an instruction parsing algorithm unit (12), a task scheduling unit (13), a path planning unit (14), a data storage unit (15), a status monitoring unit (16), and a communication transmission unit (17). The terminal humanoid robot (3) includes a robot main control unit (31), a local offline voice recognition module (32), a limb action execution module (33), a mobile drive module (34), a human-computer interaction feedback module (35), a positioning module (36), a robot communication module (37), and a robot body shell (38). The robot's outer shell (38) is made of rigid material or flexible biomimetic material; The user terminal (2) is used to send remote control commands to the cloud management platform (1) and view the device operation status. The cloud management platform (1) processes the commands through the command parsing algorithm unit (12) and the task scheduling unit (13) and then sends them to the terminal humanoid robot (3). The terminal humanoid robot (3) drives the corresponding module to perform operations through the robot main control unit (31). At the same time, the terminal humanoid robot (3) can receive voice commands offline through the local offline voice recognition module (32) and complete the command execution autonomously without relying on the cloud network.

2. The general-purpose humanoid robot cloud control system according to claim 1, characterized in that, The terminal humanoid robot (3) has dual working modes of companionship and operation, and can switch modes through cloud platform or local commands.

3. The general-purpose humanoid robot cloud control system according to claim 1, characterized in that, The local offline speech recognition module (32) has a built-in lightweight high-frequency instruction set, which only stores the feature values ​​of high-frequency speech instructions in preset scenarios. It can independently complete instruction recognition and signal transmission in the absence of a network environment.

4. The general-purpose humanoid robot cloud control system according to claim 1, characterized in that, The path planning unit (14) has a built-in multi-scene semantic map library. The semantic map adopts a coordinate + label data structure to mark scene function labels for different points. It can identify and navigate to the living area of ​​the home scene, the service point of the commercial scene, or the work station of the industrial scene, and automatically adjust the movement speed and obstacle avoidance strategy according to the scene type.

5. A general-purpose humanoid robot cloud control system according to claim 1, characterized in that, The status monitoring unit (16) receives the running status, location, power consumption and task execution data uploaded by the terminal humanoid robot (3) in real time, and automatically pushes early warning information to the user terminal (2) in abnormal conditions.

6. The general-purpose humanoid robot cloud control system according to claim 1, characterized in that, The human-computer interaction feedback module (35) has scene adaptive feedback logic: when in the caregiver mode, it calls the preset emotional voice library to generate human-like emotional feedback; when in the work mode, it calls the status prompt library to generate standardized status feedback containing task progress and fault codes.

7. A general-purpose humanoid robot cloud control system according to claim 1, characterized in that, The cloud-based management and control platform (1) is a general-purpose scheduling platform that can simultaneously connect multiple humanoid robots (3) with different scenarios and shell materials to achieve unified scheduling, group management and batch instruction issuance.

8. A general-purpose humanoid robot cloud control system according to claim 1, characterized in that, The robot communication module (37) and communication transmission unit (17) adopt any one or more wireless communication methods such as WiFi, Bluetooth, and narrowband IoT, and support seamless switching between offline operation and network data synchronization.

9. A general-purpose humanoid robot cloud control system according to claim 1, characterized in that, The flexible biomimetic materials include silicone, TPE, and soft fabrics, while the rigid materials include metals and hard engineering plastics.

10. A general-purpose cloud-based control method for humanoid robots based on the system described in any one of claims 1-9, characterized in that, Includes the following steps: (1) Users can input control commands through user terminals (2) and upload them to the cloud management platform (1), or send local voice commands directly to the terminal humanoid robot (3); (2) The cloud management platform (1) parses instructions through the instruction parsing algorithm unit (12), the path planning unit (14) plans the execution path based on the coordinate + label semantic map, and the task scheduling unit (13) issues scheduling instructions; (3) The terminal humanoid robot (3) receives cloud instructions through the robot communication module (37) or recognizes local instructions by matching a lightweight high-frequency instruction set through the local offline voice recognition module (32) and transmits them to the robot main control unit (31). (4) The robot main control unit (31) drives the movement drive module (34) and the limb action execution module (33) to perform corresponding operations, and the human-machine interaction feedback module (35) outputs corresponding feedback according to the current working mode; (5) The terminal humanoid robot (3) uploads the operation data to the cloud management platform (1) in real time, and the status monitoring unit (16) completes status monitoring and abnormal warning; (6) When the network is interrupted, the terminal humanoid robot (3) automatically switches to local offline operation mode and automatically completes data synchronization after the network is restored.

Images