Home intelligent guardian robot and intelligent detection and obstacle avoidance method thereof

By combining a symmetrical support arm, a rotatable connecting rod, a Mecanum wheel, and an AI intelligent gimbal design, the problem of mobility and functional integration of traditional home security devices in complex environments has been solved. This enables all-area, blind-spot-free detection and autonomous obstacle avoidance, enhancing the intelligence and security of home security.

CN122353526APending Publication Date: 2026-07-10
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202610673162.4
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Filing Date
2026-05-15
Publication Date
2026-07-10

AI Technical Summary

Technical Problem

Traditional home security devices cannot adapt to the complex home environment, lack mobility and functional integration, have low security recognition accuracy, imperfect early warning mechanisms, weak obstacle avoidance capabilities, cannot achieve full-area detection without blind spots, have single intelligent sensing links, and the devices are independent and cannot work together.

Method used

It adopts a combination structure of symmetrical support arms, rotatable connecting rods and Mecanum wheels, integrates AI intelligent gimbal and multi-source sensors to achieve full-domain perception and visual recognition, autonomous obstacle avoidance through multi-source obstacle acquisition and path attitude control, and intelligent detection by combining infrared scanning and voice linkage.

Benefits of technology

It enhances the intelligence and security of home security devices, enabling all-around detection without blind spots, autonomous obstacle avoidance, and accurate identification, thereby improving the responsiveness and ease of use of home security protection.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122353526A_ABST
    Figure CN122353526A_ABST
Patent Text Reader

Abstract

This invention discloses a home intelligent guardian robot and its intelligent detection and obstacle avoidance method, relating to the field of intelligent home robot technology. The robot includes two sets of symmetrically arranged support arms. The inner sides of the two sets of support arms are rotatably connected to both ends of a connecting rod, and the middle of the connecting rod is rotatably connected to the outer wall of the main unit. Mecanum wheels are fixedly mounted at the bottom end of each set of support arms. An auxiliary functional unit is fixedly arranged on the outer side of the main unit, and an AI intelligent gimbal is fixedly mounted on the outer wall of the main unit or the auxiliary functional unit. This invention features an ingenious robot design, deeply integrating mechanical structure, protection, functional expansion, and intelligent perception to adapt to the home environment. Its intelligent detection and obstacle avoidance method allows the robot to autonomously perceive, identify, and make operational decisions, enhancing home security response capabilities, ensuring smooth movement, improving intelligence, safety, and convenience, and meeting daily family needs.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention belongs to the field of smart home robot technology, specifically relating to a home intelligent guardian robot and its intelligent detection and obstacle avoidance method. Background Technology

[0002] With the rapid development of the smart home industry, the demand for intelligent home security, daily care, and environmental monitoring continues to grow. The demand for integrated, automated, and mobile intelligent security devices in home settings is becoming increasingly urgent. As an important carrier of smart home terminals, intelligent mobile robots integrate technologies such as mobile execution, environmental perception, intelligent recognition, and data interaction. They can undertake tasks such as indoor security patrols, anomaly monitoring, and emergency alarms, becoming a core product to make up for the limitations of traditional fixed security equipment. The complex spatial layout of home living environments, characterized by diverse obstacles, narrow passage areas, and scattered monitoring points, places stringent requirements on the mobility, structural stability, functional integration, and autonomous decision-making capabilities of robots. Currently, the industry urgently needs home security devices that combine all-area perception, autonomous movement, intelligent protection, and multi-functional assistance to meet the all-round, all-weather security needs of modern families.

[0003] Traditional home security devices and conventional mobile robots have many application shortcomings and cannot adapt to the actual usage requirements of home scenarios. Most security monitoring devices are fixed installations and cannot move and adjust their monitoring positions autonomously. Their coverage is limited and it is difficult to achieve full-area, blind-spot-free detection indoors. Existing mobile robots have simple structural designs and insufficient adaptability of their mobile mechanisms. When facing the complex environment of a home, their obstacle avoidance and obstacle crossing capabilities are weak, and they are prone to problems such as movement obstruction and body imbalance. The integration of device functions is low. Security identification, environmental monitoring, and auxiliary services are independent of each other and cannot achieve collaborative operation. The intelligent perception link often relies on a single sensor module, resulting in insufficient accuracy in identity recognition. The early warning mechanism is imperfect and prone to false alarms or missed alarms. Furthermore, the early warning linkage and remote interaction capabilities are lacking. The obstacle avoidance control logic is simple and cannot autonomously adjust the body posture according to obstacle characteristics. The movement stability and environmental adaptability are poor. The overall protection effect and user experience are difficult to meet the standards for home use. Summary of the Invention

[0004] The purpose of this invention is to overcome the shortcomings of existing technologies and provide a home intelligent guardian robot and its intelligent detection and obstacle avoidance method. It utilizes a combination of symmetrical support arms, rotatable connecting rods, and Mecanum wheels to form a flexible mobile structure. The main unit integrates core control and sensing components, while auxiliary functional units integrate monitoring, interaction, and storage functions. An AI intelligent gimbal enables omni-channel perception and visual recognition. The robot relies on infrared scanning, identity verification, and voice interaction to complete intelligent detection. It achieves autonomous obstacle avoidance through multi-source obstacle acquisition, folding obstacle-crossing adjustment, and path posture control. The overall structure is stable and highly adaptable to various environments, enabling it to autonomously perform tasks such as security patrols, human recognition, anomaly warnings, and environmental monitoring. This invention integrates mechanical structure optimization and intelligent control technology, overcoming the shortcomings of traditional devices such as limited mobility, single function, and weak perception. It comprehensively enhances the intelligence, security, and practicality of home guardianship, adapting to various home protection scenarios.

[0005] To solve the above-mentioned technical problems, the present invention provides the following technical solution: On the one hand, a home intelligent guardian robot includes two sets of symmetrically arranged support arms, the inner sides of the two sets of support arms are respectively rotatably connected to the two ends of a connecting rod, the middle part of the connecting rod is rotatably connected to the outer wall of the main unit, a Mecanum wheel is fixedly mounted at the bottom end of each set of support arms, an auxiliary function unit is fixedly arranged on the outer side of the main unit, and an AI intelligent gimbal is fixedly mounted on the outer wall of the main unit or the auxiliary function unit.

[0006] Furthermore, the internal cavity of the support arm is a closed hollow structure, a drive motor is fixedly installed inside the cavity of the support arm, a counterweight is fixedly placed at the bottom of the cavity of the support arm, and the output shaft of the drive motor is fixedly connected to the rotational connecting shaft between the support arm and the connecting rod, for driving the support arm to swing around the connecting rod.

[0007] Furthermore, the host includes a closed protective housing, inside which a main control chip, an AI chip, a Wi-Fi module, four ultrasonic radars, an obstacle avoidance motor, and a power module are sequentially fixed. The AI ​​chip, Wi-Fi module, four ultrasonic radars, obstacle avoidance motor, and power module are each electrically connected to the main control chip via circuits.

[0008] Furthermore, the auxiliary functional unit includes an integrated support box, the internal cavity of which is provided with a transverse sliding guide rail, the surface of which is fitted with a smart medicine box, the upper surface of which is fitted with an OLED touch screen, and the side walls of which are uniformly provided with regularly arranged sensor interfaces.

[0009] Furthermore, the internal slots of the sensor interface are respectively fixedly equipped with smoke sensing elements, gas detection elements, air monitoring elements and temperature and humidity acquisition elements, and each type of sensing element is connected to the main control chip through a circuit; the AI ​​smart gimbal includes an outer shell, and an OpenMV gimbal camera is fixedly installed inside the outer shell, and an infrared human body sensor is fixedly attached to the top surface of the OpenMV gimbal camera.

[0010] On the other hand, a smart detection method for a home smart guardian robot, the specific steps of which are as follows: Full-area infrared scanning: Relying on the infrared human body sensor mounted on the AI ​​intelligent gimbal, it continuously detects changes in infrared radiation in the environment and completes continuous detection operations in indoor spaces. Visual identity verification: After the infrared human body sensor captures human activity signals, the OpenMV gimbal camera is activated to simultaneously perform image acquisition, human body tracking and face comparison operations, and runs a stranger identification algorithm. When the face comparison is abnormal, an early warning signal is triggered and output. Command-linked control: Continuously collects environmental voice signals, completes the matching and comparison of preset voice commands, and runs the warning cancellation execution algorithm simultaneously. The warning signal is terminated after the voice command matching is completed. When the voice command matching is missing, a remote alarm text signal is transmitted to the outside.

[0011] Furthermore, in the aforementioned visual identity verification, the mathematical expression for the algorithm for determining if a person is a stranger is: ;in The overall judgment coefficient for strangers. Here, p represents the fixed weighting coefficient for infrared human detection, and p is the real-time detection value of the infrared human sensor. Here, M represents the fixed weighting coefficients for facial feature matching, and M is the facial feature comparison value collected by the OpenMV gimbal camera.

[0012] Furthermore, in the aforementioned instruction linkage control, the mathematical expression for the warning cancellation execution algorithm is: ;in For remote signal transmission coefficient, A fixed weighting coefficient is used to determine the warning duration. This represents the real-time percentage of the warning time. For speech recognition, fixed weight coefficients are used. Match numerical values ​​to on-site voice commands.

[0013] On one hand, there is an intelligent obstacle avoidance method for a home smart guardian robot, the specific steps of which are as follows: Multi-source obstacle acquisition: Relying on four-sided ultrasonic radar and bottom contact sensor to synchronously collect environmental data, and run obstacle crossing judgment algorithm based on the collected data to complete the comparison and judgment of obstacle attributes; Folding obstacle crossing adjustment: When the obstacle properties meet the obstacle crossing conditions, the control connecting rod rotates and swings to adjust the spatial attitude of the main unit and complete the segmented crossing movement of the machine body; Path attitude control: When the obstacle attributes do not meet the obstacle crossing conditions, the Mecanum wheel travel trajectory is adjusted to complete the path change. After the crossing action is completed, the attitude reset algorithm is run to control the connecting rod and support arm to rotate and reset in an orderly manner.

[0014] Furthermore, in the multi-source obstacle acquisition, the mathematical expression of the obstacle crossing determination algorithm is: ;in For cross-obstacle comprehensive discrimination values, , , A fixed coefficient is preset for the robot structure, where h is the actual measured height of the obstacle, l is the limit travel of the connecting rod, and d is the real-time distance between the robot and the obstacle.

[0015] Compared with the prior art, the present invention has the following beneficial technical effects: (1) This invention, through the combined structural design of symmetrical support arms, rotatable connecting rods and Mecanum wheels, endows the robot with flexible mobility and variable body posture. The closed hollow layout of the support arms, combined with the built-in drive and counterweight structure, effectively optimizes the center of gravity of the whole machine, improves the stability during walking and posture adjustment, and avoids imbalance problems during operation. The main unit adopts a closed protective shell, and the internal core components are independently electrically connected to ensure reliable circuit operation and good protection. The auxiliary function unit integrates storage, touch display and multiple types of sensor interfaces, and can be equipped with a variety of environmental monitoring components. The AI ​​intelligent gimbal integrates infrared sensing and visual acquisition components to achieve all-round spatial detection and human body recognition. This design deeply integrates mechanical structure, protective performance, functional expansion and intelligent sensing, adapts to the complex indoor environment of the home, enriches the practical functions of the robot, and improves the adaptability and reliability of home protection.

[0016] (2) This invention uses intelligent detection logic of full-domain infrared scanning, visual identity recognition and command linkage control, combined with intelligent obstacle avoidance process of multi-source obstacle acquisition, folding obstacle crossing adjustment and path posture control, to enable the robot to achieve autonomous perception, accurate recognition and autonomous decision-making operation. Infrared perception and visual acquisition work together to quickly complete human body detection, identity comparison and abnormal warning. Voice command linkage can efficiently remove warnings or trigger remote alarms, enhance the response capability of home security protection. Multi-source sensors collect obstacle information synchronously, and combined with body structure adjustment, segmented obstacle crossing or path adjustment can be realized. After the action is completed, it automatically resets to ensure smooth movement. The whole method does not require continuous manual operation. It autonomously completes environmental monitoring, safety alarm, intelligent obstacle avoidance and autonomous movement, significantly improving the intelligence level, security and convenience of home intelligent protection, and effectively meeting the daily care and security protection needs of the family.

[0017] Other advantages, objectives and features of the invention will be set forth in part in the description which follows, and in part will be apparent to those skilled in the art from the following examination or study, or may be learned from the practice of the invention. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are merely some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without any creative effort.

[0019] Figure 1 This is a schematic diagram of the overall structure of the home guardian robot of the present invention; Figure 2 This is another structural schematic diagram of the home guardian robot of the present invention; Figure 3 This is another structural schematic diagram of the home guardian robot of the present invention; Figure 4 This is a schematic diagram of the hardware structure of the home guardian robot of the present invention; Figure 5 This is a flowchart illustrating the intelligent detection method for the home guardian robot of the present invention; Figure 6 This is a flowchart illustrating the intelligent obstacle avoidance method of the home guardian robot of the present invention; The diagram shows: 1. Support arm; 2. Connecting rod; 3. Main unit; 4. Mecanum wheel; 5. Auxiliary function unit; 6. AI intelligent gimbal; 3.1 Housing; 3.2 Main control chip; 3.3 AI chip; 3.4 Wi-Fi module; 3.5 Four-sided ultrasonic radar; 3.6 Obstacle avoidance motor; 3.7 Power module; 5.1 Support box; 5.2 Smart medicine box; 5.3 OLED touch display screen; 5.4 Sensor interface; 6.1 External housing; 6.2 OpenMV gimbal camera; 6.3 Infrared human body sensor. Detailed Implementation

[0020] To further illustrate the technical means and effects of the present invention in achieving its intended purpose, the following detailed description of the specific implementation methods, structures, features, and effects of the present invention, in conjunction with the accompanying drawings and preferred embodiments, is provided below.

[0021] Example 1 : Please see Figures 1 to 4 As shown in the figure, this invention provides a home intelligent guardian robot. The robot adopts a symmetrical, swingable support structure and an omnidirectional movement mechanism, integrating intelligent visual recognition, multi-dimensional environmental monitoring, autonomous obstacle avoidance, an emergency medicine box, and remote alarm functions. It can achieve all-weather, automated, and intelligent safety protection in home scenarios. Its specific structure and working method are as follows: Overall mechanical structure of the robot: In this embodiment, the core skeleton of the home intelligent guardian robot includes two sets of symmetrically arranged support arms 1. The two sets of support arms 1 are symmetrically distributed left and right around the robot host 3, ensuring the balance of the entire robot's movement and posture adjustment. The inner end faces of the two sets of support arms 1 are rotatably connected to the left and right ends of the connecting rod 2, respectively. The middle position of the connecting rod 2 is rotatably connected to the outer wall of the host 3, forming a three-section rotatable linkage structure of double support arms, connecting rod, and host. The spatial posture of the host 3 can be flexibly adjusted by the relative swinging of the support arms 1 and the connecting rod 2.

[0022] Each set of support arms 1 is fixedly equipped with a Mecanum wheel 4 at its bottom end. The Mecanum wheel 4 consists of multiple rollers distributed at a 45° angle between a single hub and the axis, supporting 360° rotation in place, lateral translation, and diagonal movement of the robot, breaking through the limitations of traditional wheeled robot movement. An auxiliary function unit 5 is fixedly deployed on the outside of the main unit 3 to integrate environmental monitoring, emergency interaction, and medicine storage functions; an AI intelligent gimbal 6 is fixedly mounted on the outer wall of the main unit 3 or the outer wall of the auxiliary function unit 5 to realize human perception, visual tracking, and identity recognition.

[0023] Internal structure and drive mechanism of the support arm: In this embodiment, the support arm 1 adopts a closed hollow structure cavity. The closed design can effectively block dust and moisture in the home environment, protect the internal components, and extend their service life. A drive motor is fixedly installed inside the cavity of the support arm 1, and a counterweight is fixedly placed at the bottom of the cavity. The counterweight can lower the center of gravity of the whole machine, improve the stability of the robot during movement, posture adjustment, and obstacle crossing, and prevent tipping.

[0024] The output shaft of the drive motor is fixedly connected to the rotating connecting shaft between the support arm 1 and the connecting rod 2. When the drive motor is powered on, it can directly drive the support arm 1 to swing around the connecting rod 2 in a directional manner. In conjunction with the rotating pair between the connecting rod 2 and the main unit 3, the whole machine can complete actions such as folding, lifting, and crossing without the need for an additional transmission mechanism. It has a fast response speed and high transmission efficiency.

[0025] The internal module composition and connection relationship of the host: The main unit 3, serving as the robot's control core, includes a closed protective shell 3.1. The protective shell 3.1 is made of ABS engineering plastic, possessing characteristics of impact resistance, wear resistance, and good insulation, making it suitable for complex home environments. Inside the protective shell 3.1, the main control chip 3.2, AI chip 3.3, Wi-Fi module 3.4, four-sided ultrasonic radar 3.5, obstacle avoidance motor 3.6, and power module 3.7 are sequentially arranged according to their functional zones.

[0026] All of the above functional modules are electrically connected to the main control chip 3.2 via independent lines, ensuring that they do not interfere with each other and communicate independently, thus guaranteeing the stability of signal transmission. Main control chip 3.2: As the core of the whole machine control, it is responsible for receiving sensor data, issuing motion commands, and coordinating the collaborative work of various modules; AI Chip 3.3: Responsible for running complex algorithms such as facial recognition, stranger detection, and intelligent obstacle avoidance to achieve edge intelligent computing; Wi-Fi module 3.4: Enables remote data transmission, mobile app integration, and smart home device control; Four-sided ultrasonic radar 3.5: Detects obstacles in all directions around the robot and acquires distance and height data; Obstacle avoidance motor 3.6: Receives instructions from main control chip 3.2 and executes obstacle avoidance actions such as support arm swing and path adjustment; Power module 3.7: Provides stable power to all electronic components of the device, ensuring battery life and operational reliability.

[0027] Auxiliary functional unit structure design: The auxiliary function unit 5 adopts an integrated molded support box 5.1, which is compact, firmly installed, and seamless, making it suitable for home use. The internal cavity of the support box 5.1 is equipped with a horizontal sliding guide rail, on which the smart medicine box 5.2 is fitted. The smart medicine box 5.2 can automatically slide out / retract along the guide rail to store emergency medicines and first aid tools, meeting the family's emergency needs.

[0028] The upper surface of the supporting cabinet 5.1 is fitted with an OLED touch screen 5.3, which adopts self-emissive technology, has low power consumption, high definition, and good color saturation. It can display temperature and humidity, air quality, alarm information, and robot status in real time, and supports touch control and voice interaction to achieve convenient human-machine interaction.

[0029] The side walls of the support housing 5.1 are evenly provided with neatly arranged sensor interfaces 5.4. The standardized interface design facilitates sensor disassembly, maintenance and upgrades, and provides an installation carrier for various types of environmental monitoring components.

[0030] Details of environmental sensing elements and AI smart gimbal: The internal slots of sensor interface 5.4 are respectively fixed to the smoke sensing element, gas detection element, air monitoring element and temperature and humidity acquisition element. The four types of sensing elements are connected to the main control chip 3.2 through dedicated lines to collect indoor smoke concentration, combustible gas leakage, PM2.5 / formaldehyde / TVOC air quality and ambient temperature and humidity data in real time. Once the data exceeds the standard, local alarm and remote notification will be triggered immediately.

[0031] The AI ​​Smart Gimbal 6 includes an outer shell 6.1, inside which an OpenMV gimbal camera 6.2 is fixedly installed. The OpenMV gimbal camera 6.2 features low power consumption, high frame rate, contour recognition, and human body tracking. An infrared human body sensor 6.3 is fixedly attached to the top surface of the OpenMV gimbal camera 6.2. The infrared human body sensor 6.3 is rigidly connected to the OpenMV gimbal camera 6.2 and synchronously fixed inside the AI ​​Smart Gimbal 6. It can rotate and adjust its angle synchronously with the gimbal, achieving full-area human body detection and visual acquisition without blind spots.

[0032] Example 2 : This embodiment provides an intelligent detection method for a home smart guardian robot, the specific functions and principles of which are as follows: Please see Figure 5 As shown, embodiments of the present invention propose an intelligent detection method for a home intelligent guardian robot. This method is adapted to a home intelligent guardian robot comprising two symmetrically arranged support arms 1, connecting rods 2, a main unit 3, Mecanum wheels 4, auxiliary function units 5, and an AI intelligent gimbal 6. It adopts a three-level collaborative processing flow of full-domain infrared scanning, visual identity recognition, and command linkage control. Relying on the multi-sensor fusion perception capability of the AI ​​intelligent gimbal 6, and in conjunction with the computing power of the main control chip 3.2 and AI chip 3.3 inside the main unit 3, it can complete continuous detection of human activity in indoor spaces, accurate identification of family members and strangers, and remote alarm intelligent control, achieving high sensitivity and low false alarm rate all-weather intelligent guardian detection in home scenarios.

[0033] The method described in this embodiment specifically includes: Full-area infrared scanning: Relying on the infrared human body sensor 6.3 mounted on the AI ​​intelligent gimbal 6, it continuously detects changes in infrared radiation in the environment and completes continuous detection operations in indoor spaces; Visual identity verification: After the infrared human body sensor 6.3 captures human activity signals, it starts the OpenMV gimbal camera 6.2 to simultaneously perform image acquisition, human body tracking and face comparison operations, and runs a stranger detection algorithm. When the face comparison is abnormal, an early warning signal is triggered and output. Command-linked control: Continuously collects environmental voice signals, completes the matching and comparison of preset voice commands, and runs the warning cancellation execution algorithm simultaneously. After the voice command matching is completed, the warning signal is terminated. When the voice command matching is missing, the remote alarm text signal is transmitted to the outside through the Wi-Fi module 3.4 of the host 3.

[0034] Full-area infrared scanning: In the full-area infrared scanning step, the infrared human body sensor 6.3 is locked and fixed to the top of the OpenMV gimbal camera 6.2. The infrared human body sensor 6.3 and the OpenMV gimbal camera 6.2 are synchronously fixed inside the outer shell 6.1 of the AI ​​intelligent gimbal 6, maintaining synchronous rotation and angle adjustment. The AI ​​intelligent gimbal 6 is fixedly mounted on the outer wall of the main unit 3 or the auxiliary functional unit 5, enabling full-area infrared detection without blind spots indoors. The AI ​​intelligent gimbal 6 drives the infrared human body sensor 6.3 to perform cyclic scanning of the indoor space according to a preset cycle. The infrared human body sensor 6.3 transmits infrared radiation detection data to the main control chip 3.2 inside the main unit 3 in real time. The main control chip 3.2 analyzes the detection data in real time. When a human infrared radiation signal is detected, it immediately sends a visual identification start command to the AI ​​chip 3.3, simultaneously triggering the subsequent visual identity identification process.

[0035] Visual identity verification: In the visual identity verification step, after receiving the start command from the main control chip 3.2, the AI ​​chip 3.3 controls the OpenMV gimbal camera 6.2 inside the AI ​​smart gimbal 6 to start. The OpenMV gimbal camera 6.2 collects indoor image data in real time, simultaneously completing human dynamic tracking and facial feature extraction operations, and comparing the extracted facial features with the locally pre-stored family member facial feature database to generate a facial feature comparison value M. The main control chip 3.2 simultaneously runs a stranger detection algorithm, the mathematical expression of which is: ;in The overall judgment coefficient for strangers. Here, a fixed weighting coefficient is used for infrared human body detection, and p is the real-time detection value of the infrared human body sensor (6.3). The fixed weight coefficients for facial feature matching are given, and M is the facial feature comparison value collected by the OpenMV gimbal camera (6.2). All of the above parameters are fixed parameters preset before the robot leaves the factory and real-time acquisition parameters; no additional parameters are added. When the comprehensive stranger judgment coefficient Φ reaches the warning threshold, the main control chip 3.2 immediately generates a warning signal and transmits it to the OLED touch screen 5.3 of the auxiliary function unit 5 for local notification.

[0036] For example, in the visual identity recognition process, a human stranger judgment algorithm is used for calculation. The inherent parameters preset by the robot and collected in real time are substituted to complete the calculation of the stranger comprehensive judgment coefficient. The main control chip 3.2 determines whether to trigger the warning signal based on the calculation result. The whole process only uses the inherent parameters in the algorithm expression, without adding any additional parameters.

[0037] Command-based linkage control: In the aforementioned instruction-linked control steps, after the main control chip 3.2 outputs a warning signal, it simultaneously activates the voice acquisition module to continuously pick up indoor environmental voice signals and transmits them to the AI ​​chip 3.3 for voice recognition. The AI ​​chip 3.3 matches the recognition result with a preset warning cancellation voice command library to generate a matching value K for the on-site voice command. The main control chip 3.2 simultaneously runs the warning cancellation execution algorithm, the mathematical expression of which is: ;in For remote signal transmission coefficient, A fixed weighting coefficient is used to determine the warning duration. This represents the real-time percentage of the warning time. For speech recognition, fixed weight coefficients are used. The values ​​are matched to the on-site voice commands. All parameters mentioned above are preset fixed parameters and real-time collected parameters for the robot; no additional parameters are added. If the voice command matching is successful and the remote signal transmission coefficient Λ meets the release condition, the main control chip 3.2 immediately terminates the warning signal; if the voice command matching is missing but the remote signal transmission coefficient Λ meets the transmission condition, the main control chip 3.2 transmits the remote alarm text signal externally through the Wi-Fi module 3.4 of the host 3.

[0038] For example, during the command linkage control process, the early warning cancellation execution algorithm is used to calculate the inherent parameters preset by the robot and collected in real time to complete the calculation of the remote signal transmission coefficient. The main control chip 3.2 determines whether to terminate the early warning or transmit the alarm signal based on the calculation result. The entire process only uses the inherent parameters in the algorithm expression, without adding any additional parameters.

[0039] Example 3 : This embodiment provides an intelligent obstacle avoidance method for a home smart guardian robot, the specific functions and principles of which are as follows: Please see Figure 6As shown, embodiments of the present invention propose an intelligent obstacle avoidance method for a home intelligent guardian robot. This method is adapted to a home intelligent guardian robot comprising two symmetrically arranged support arms 1, connecting rods 2, a main unit 3, Mecanum wheels 4, auxiliary function units 5, and an AI intelligent gimbal 6. It adopts a hierarchical processing logic of multi-source obstacle acquisition, folding obstacle crossing adjustment, and path posture control. It can complete obstacle attribute identification, segmented obstacle crossing, and automatic posture reset by using multi-source data fusion from the four ultrasonic radars 3.5 inside the main unit 3 and the bottom contact sensor, combined with the coordinated action of the obstacle avoidance motor 3.6 and the drive motor inside the support arm 1. This allows for autonomous obstacle avoidance and stable movement in various complex indoor environments.

[0040] The method described in this embodiment specifically includes: Multi-source obstacle acquisition: Relying on the four ultrasonic radars 3.5 inside the host 3 and the bottom contact sensor to synchronously collect environmental data, the obstacle crossing judgment algorithm is run based on the collected data to complete the comparison and judgment of obstacle attributes; Folding obstacle crossing adjustment: When the obstacle properties meet the obstacle crossing conditions, the main control chip 3.2 controls the connecting rod 2 to rotate and swing, adjust the spatial posture of the main unit 3, and complete the segmented crossing movement of the machine body; Path attitude control: When the obstacle attributes do not meet the obstacle crossing conditions, the main control chip 3.2 adjusts the travel trajectory of the Mecanum wheel 4 to complete the path change. After the crossing action is completed, the attitude reset algorithm is run to control the connecting rod 2 and the support arm 1 to rotate and reset in an orderly manner.

[0041] Multi-source obstacle acquisition: In the multi-source obstacle acquisition step, four ultrasonic radars 3.5 are fixedly installed inside the protective shell 3.1 of the main unit 3. These four ultrasonic radars 3.5 detect obstacles in the front, back, left, and right directions of the robot, respectively, and collect real-time data on obstacle height and the distance between the robot and the obstacle. Simultaneously, the bottom contact sensor collects ground contact status and obstacle contact data. Both types of sensor data are independently transmitted to the main control chip 3.2 via lines. After integrating and analyzing the multi-source data, the main control chip 3.2 runs the obstacle crossing determination algorithm. The mathematical expression of the obstacle crossing determination algorithm is: ;in For cross-obstacle comprehensive discrimination values, , , A fixed coefficient is preset for the robot structure, h is the actual measured height of the obstacle, l is the limit travel value of the connecting rod (2), and d is the real-time distance between the robot and the obstacle. The above parameters are all inherent preset parameters of the robot structure and real-time data collected by the sensors, with no additional parameters added. The main control chip 3.2 determines whether the obstacle meets the obstacle crossing conditions based on the obstacle crossing comprehensive judgment value Γ, and then selects to execute the subsequent folding obstacle crossing adjustment or path posture control process.

[0042] For example, during the multi-source obstacle acquisition process, an obstacle crossing judgment algorithm is used for calculation. The inherent parameters of the robot structure preset and the real-time acquisition of the sensors are substituted to complete the calculation of the comprehensive obstacle crossing judgment value. The main control chip 3.2 determines whether the obstacle meets the obstacle crossing conditions based on the calculation result. The entire process only uses the inherent parameters in the algorithm expression, without adding any additional parameters.

[0043] Folding obstacle crossing adjustment: In the folding obstacle-crossing adjustment step, when the obstacle properties meet the obstacle-crossing conditions, the main control chip 3.2 sends a control command to the drive motor inside the support arm 1. The support arm 1 is a closed hollow structure, and the output shaft of the drive motor inside its cavity is fixedly connected to the rotational connecting shaft between the support arm 1 and the connecting rod 2. The drive motor drives the support arm 1 to swing around the connecting rod 2, and the middle part of the connecting rod 2 rotates synchronously around the outer wall of the main unit 3. Through the coordinated swinging of the support arm 1 and the connecting rod 2, the height and spatial posture of the main unit 3 are adjusted. With the micro-motion of the Mecanum wheel 4 at the bottom of the support arm 1, the robot body can cross obstacles in segments. During the crossing process, the four ultrasonic radars 3.5 continuously detect the surrounding environment to avoid collisions between the robot body and obstacles, ensuring the operational safety of the main unit 3, the auxiliary function unit 5, and the AI ​​intelligent gimbal 6.

[0044] Path attitude control: In the path attitude control step, when the obstacle attributes do not meet the obstacle-crossing conditions, the main control chip 3.2 sends control commands to the Mecanum wheel 4 drive assembly. Utilizing the omnidirectional movement characteristics of the Mecanum wheel 4, the robot's direction, speed, and trajectory are adjusted to avoid the obstacle and complete the path change. After the robot completes its obstacle-crossing or obstacle-avoiding action, the main control chip 3.2 runs the attitude reset algorithm. The mathematical expression of the attitude reset algorithm is: ;in Values ​​for adjusting the body's attitude. , Preset fixed coefficients for motor drive. g is the real-time rotation angle value of the connecting rod (2), and g is the real-time feedback value of the bottom contact sensor. The above parameters are the comprehensive obstacle crossing judgment values. All of the above parameters are the preset fixed parameters of the motor, the real-time data collected by the sensor, and the inherent parameters of the previous calculation. There are no additional parameters. The main control chip 3.2 adjusts the value Ω according to the robot's posture and controls the internal drive motor and obstacle avoidance motor 3.6 of the support arm 1 to rotate in opposite directions, driving the connecting rod 2 and the support arm 1 to rotate in an orderly manner until the robot returns to its initial horizontal posture, completing the overall posture reset.

[0045] For example, during the path attitude control and attitude reset process, the attitude reset algorithm is used for calculation. The inherent parameters of the motor preset, the sensor real-time acquisition and the previous calculation are substituted to complete the calculation of the body attitude adjustment value. The main control chip 3.2 controls the connecting rod 2 and the support arm 1 to reset according to the calculation result. The whole process only uses the inherent parameters in the algorithm expression, without adding any additional parameters.

[0046] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any simple modifications, equivalent changes and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. A home intelligent guardian robot, characterized in that, It includes two sets of symmetrically arranged support arms (1), the inner side of the two sets of support arms (1) are respectively rotatably connected to the two ends of the connecting rod (2), the middle part of the connecting rod (2) is rotatably connected to the outer wall of the host (3), the bottom end of each set of support arms (1) is fixedly equipped with a Mecanum wheel (4), the outer side of the host (3) is fixedly arranged with an auxiliary function unit (5), and an AI smart gimbal (6) is fixedly equipped on the outer wall of the host (3) or the auxiliary function unit (5).

2. The home intelligent guardian robot according to claim 1, characterized in that, The internal cavity of the support arm (1) is a closed hollow structure. A drive motor is fixedly installed inside the cavity of the support arm (1). A counterweight is fixedly placed at the bottom of the cavity of the support arm (1). The output shaft of the drive motor is fixedly connected to the rotating connecting shaft between the support arm (1) and the connecting rod (2) to drive the support arm (1) to swing around the connecting rod (2).

3. The home intelligent guardian robot according to claim 1, characterized in that, The host (3) includes a closed protective shell (3.1). Inside the protective shell (3.1), a main control chip (3.2), an AI chip (3.3), a Wi-Fi module (3.4), a four-sided ultrasonic radar (3.5), an obstacle avoidance motor (3.6), and a power module (3.7) are sequentially fixed. The AI ​​chip (3.3), Wi-Fi module (3.4), four-sided ultrasonic radar (3.5), obstacle avoidance motor (3.6), and power module (3.7) are each electrically connected to the main control chip (3.2) through lines.

4. A home intelligent guardian robot according to claim 1, characterized in that, The auxiliary function unit (5) includes an integrated support box (5.1). The internal cavity of the support box (5.1) is provided with a transverse sliding guide rail. The surface of the sliding guide rail is fitted with an intelligent medicine box (5.2). The upper surface of the support box (5.1) is fitted with an OLED touch screen (5.3). The side wall of the support box (5.1) is uniformly provided with regularly arranged sensor interfaces (5.4).

5. A home intelligent guardian robot according to claim 1, characterized in that, The internal slots of the sensor interface (5.4) are respectively fixedly equipped with smoke sensing elements, gas detection elements, air monitoring elements and temperature and humidity acquisition elements. The various sensing elements are respectively connected to the main control chip (3.2) through lines. The AI ​​smart gimbal (6) includes an outer shell (6.1). An OpenMV gimbal camera (6.2) is fixedly installed inside the outer shell (6.1). An infrared human body sensor (6.3) is fixedly attached to the top end face of the OpenMV gimbal camera (6.2).

6. A method for intelligent detection of a home intelligent guardian robot, applicable to the home intelligent guardian robot described in any one of claims 1-5, characterized in that, The specific steps of this method are as follows: Full-area infrared scanning: Relying on the infrared human body sensor (6.3) mounted on the AI ​​intelligent gimbal (6), it continuously detects changes in infrared radiation in the environment and completes continuous detection operations in the indoor space; Visual identity verification: After the infrared human body sensor (6.3) captures human activity signals, it starts the OpenMV gimbal camera (6.2) to simultaneously perform image acquisition, human body tracking and face comparison operations, and runs the human stranger judgment algorithm. When the face comparison is abnormal, it triggers the output of the warning signal. Command-linked control: Continuously collects environmental voice signals, completes the matching and comparison of preset voice commands, and runs the warning cancellation execution algorithm simultaneously. The warning signal is terminated after the voice command matching is completed. When the voice command matching is missing, a remote alarm text signal is transmitted to the outside.

7. The intelligent detection method for a home intelligent guardian robot according to claim 6, characterized in that, In the visual identity verification process, the mathematical expression for the algorithm for determining if a person is a stranger is: ;in The overall judgment coefficient for strangers. Here, a fixed weighting coefficient is used for infrared human body detection, and p is the real-time detection value of the infrared human body sensor (6.3). M is a fixed weight coefficient for face feature matching, and M is the face feature comparison value collected by the OpenMV gimbal camera (6.2).

8. The intelligent detection method for a home intelligent guardian robot according to claim 6, characterized in that, In the aforementioned instruction linkage control, the mathematical expression for the warning cancellation execution algorithm is: ;in For remote signal transmission coefficient, A fixed weighting coefficient is used to determine the warning duration. This represents the real-time percentage of the warning time. For speech recognition, fixed weight coefficients are used. Match numerical values ​​to on-site voice commands.

9. A method for intelligent obstacle avoidance in a home intelligent guardian robot, applicable to the home intelligent guardian robot described in any one of claims 1-5, characterized in that, The specific steps of this method are as follows: Multi-source obstacle acquisition: Relying on four ultrasonic radars (3.5) and bottom contact sensors to synchronously acquire environmental data, and run obstacle crossing judgment algorithm based on the acquired data to complete the comparison and discrimination of obstacle attributes; Folding obstacle crossing adjustment: When the obstacle properties meet the obstacle crossing conditions, control the connecting rod (2) to rotate and swing, adjust the spatial attitude of the host (3), and complete the segmented crossing movement of the body; Path attitude Control: When the obstacle attributes do not meet the obstacle crossing conditions, adjust the travel trajectory of the Mecanum wheel (4) to complete the path change. After the crossing action is completed, run the posture reset algorithm to control the connecting rod (2) and the support arm (1) to rotate and reset in an orderly manner.

10. The intelligent obstacle avoidance method for a home intelligent guardian robot according to claim 9, characterized in that, In the multi-source obstacle acquisition, the mathematical expression of the obstacle crossing determination algorithm is: ;in For cross-obstacle comprehensive discrimination values, , , A fixed coefficient is preset for the robot structure, h is the actual measured height of the obstacle, l is the limit travel value of the connecting rod (2), and d is the real-time distance between the robot and the obstacle.