Universal science and technology information service robot based on AI artificial intelligence

Abstract

The invention discloses a generic science and technology information service robot based on AI artificial intelligence, which comprises a robot main body, the robot main body comprises hardware equipment and a software system, and the hardware equipment comprises a visualization assembly, a placement assembly, a multi-sensor fusion assembly and a moving assembly; the software system comprises a multi-sensor fusion module, a man-machine interaction module, a visual information display module and a processing module. The high-precision microphone array is carried on the robot body and used for voice recognition, the emotion state of a user can be judged by analyzing the tone, speed, volume change and other characteristics of voice, meanwhile, the visual camera is arranged, the facial expression recognition technology and the gesture analysis technology are introduced, and the user experience is improved. The deep learning model is used for recognizing facial muscle movement of the user, accurately judging emotions such as joy, anger, sorrow and joy, multi-mode emotion information fusion perception is achieved, an instruction is given to the robot through gestures of the user when the environment is noisy, and man-machine interaction of the robot is improved.

Description

Technical Field

[0001] The present invention relates to a robot device, in particular to a pan-tech information service robot based on AI artificial intelligence, belonging to the technical field of robot equipment. Background Art

[0002] Service robots are a young member of the robot family. So far, there is no strict definition. Different countries have different understandings of service robots. They can be divided into professional service robots and personal / family service robots. Service robots have a wide range of applications and are mainly engaged in maintenance, repair, transportation, cleaning, security, rescue, monitoring and other tasks.

[0003] After searching, Chinese patent number CN110815256A discloses a pan-tech information service robot based on AI artificial intelligence. Two thermal sensors are symmetrically installed on the four side walls of the control box. The thermal sensors can capture the traces of people or thermal sensing equipment in the room, and according to the preset program, move to the position of the thermal detection object to work, while the ultrasonic sensor can avoid obstacles through detection, and the microphone can be set up for program control through voice recognition. However, the factory environment is complex. When using the information service robot, due to the large number of workers and high environmental noise, when issuing instructions to the information service robot through the microphone, the surrounding environment is relatively noisy, and the robot has problems of being unable to recognize or unclear recognition, resulting in slow operation and affecting work progress. Summary of the invention

[0004] The purpose of the present invention is to provide a pan-tech information service robot based on AI artificial intelligence in order to solve the above problems.

[0005] The present invention achieves the above-mentioned purpose through the following technical solutions: a pan-tech information service robot based on AI artificial intelligence, including a robot body, the robot body including hardware equipment and a software system, the hardware equipment including a visualization component, a placement component, a multi-sensor fusion component and a mobile component;

[0006] The software system includes a multi-sensor fusion module, a human-computer interaction module, a visual information display module and a processing module;

[0007] The multi-sensor fusion module is used to enable the robot body to accurately perceive the overall environment and enable the robot body to make intelligent decisions. The data collected by the multi-sensor fusion component is processed and then displayed through the visual information display module. The multi-sensor fusion module includes a data acquisition submodule and a data fusion processing submodule. The data acquisition submodule includes a laser radar acquisition unit, a visual camera acquisition unit, a thermal sensor acquisition unit, and an ultrasonic sensor acquisition unit.

[0008] The human-computer interaction module is used to improve the usage needs of customers. The human-computer interaction module includes a voice interaction submodule, a gesture interaction submodule, a touch screen interaction submodule, a visual interaction submodule and a micro-expression interaction submodule;

[0009] The visual information display module is used to present abstract knowledge in a concrete way, and can also perform abnormal warning and troubleshooting. The visual information display module can enhance the experience of the human-computer interaction module.

[0010] Preferably, the visualization component is used to improve the interactivity of the user, and the visualization component includes a vertical plate, a support plate, a display host, an expression display screen and a touch screen. The vertical plate is provided on one side of the robot body, the support plate is provided on the top of the vertical plate, the display host is provided on the top of the support plate, the touch screen is provided on the side of the vertical plate away from the robot body, and the expression display screen is provided on the side of the display host close to the touch screen;

[0011] The moving assembly is used for the robot body to move and rotate at multiple angles during use. The moving assembly includes a base, universal wheels, a drive motor and a placement block. The bottom end of the robot body is provided with a base, the bottom end of the base and the four corners are provided with universal wheels, the middle of the top end of the base is provided with a drive motor, and the execution end of the drive motor is connected to the placement block;

[0012] The placement component is used for users to place some items on the robot body for transportation during use. The placement component includes a vertical plate, an inclined plate, a limit rod and a placement frame. The vertical plate is provided on the side of the vertical plate away from the touch display screen and located at the top of the placement block. The bottom end of the vertical plate is connected to the inclined plate. The top end of the placement block is provided with a limit rod on the side away from the inclined plate. The limit rod is provided with a plurality of placement frames in sequence from top to bottom.

[0013] The multi-sensor fusion component cooperates with the visualization component to improve the user's interactivity, and at the same time enables the robot body to better judge the environment, avoid risks and issue early warnings. The multi-sensor fusion component includes a laser radar, a visual camera, a thermal sensor and an ultrasonic sensor. A laser radar is provided on the top of the display host, a visual camera is provided on the top of the touch display screen, thermal sensors are provided on the four sides of the placement block and on both sides of the display host, and a laser radar is provided around the bottom of the base near the position of the universal wheel.

[0014] Preferably, the voice interaction submodule is based on a voice recognition algorithm and a voice synthesis algorithm, the gesture interaction submodule is based on a gesture recognition algorithm and a gesture interaction feedback algorithm, the touch screen interaction submodule is based on a touch operation response algorithm and a task progress display algorithm, the visual interaction submodule is based on an expression animation generation algorithm and a knowledge visualization display algorithm, and the micro-expression interaction submodule is based on a micro-expression capture algorithm, a micro-expression recognition algorithm and a micro-expression interaction application algorithm.

[0015] Preferably, the laser radar acquisition unit is based on a laser emission and reception control algorithm and a point cloud data construction algorithm; the visual camera acquisition unit is based on an automatic focus and exposure control algorithm and a data compression algorithm; the thermal sensor acquisition unit is based on a temperature calculation and compensation algorithm, a filtering and denoising algorithm, an image reconstruction and enhancement algorithm, and a target recognition and tracking algorithm; and the ultrasonic sensor acquisition unit is based on a transit time algorithm and an algorithm for taking an average value of multiple measurements.

[0016] Preferably, the laser radar acquisition unit, the visual camera acquisition unit, the thermal sensor acquisition unit and the ultrasonic sensor acquisition unit in the multi-sensor fusion module start data acquisition work synchronously.

[0017] Preferably, the support plate is arranged at the top ends of the vertical plate and the inclined plate, and the support plate, the vertical plate and the inclined plate are arranged in a triangular shape.

[0018] Preferably, speakers are provided inside the vertical plate and the inclined plate, and a microphone is provided on the side of the touch display screen.

[0019] Preferably, a side of the placement frame away from the limiting rod is fixedly connected to the inclined plate.

[0020] Preferably, the base and the placement block are arranged in a rectangular shape, and the four corners of the base and the placement block are arranged in an arc shape.

[0021] The present invention has the following beneficial effects:

[0022] 1. The robot is equipped with a high-precision microphone array, which is not only used for voice recognition, but also can judge the user's emotional state by analyzing the voice's tone, speaking speed, volume changes and other characteristics. At the same time, it is equipped with a visual camera, introduces facial expression recognition technology and gesture analysis technology, uses a deep learning model to recognize the user's facial muscle movements, accurately judges emotions such as joy, anger, sorrow, and happiness, and realizes multi-modal emotional information fusion perception. In a noisy environment, the robot can be instructed by the user's gestures to improve the robot's human-computer interaction;

[0023] 2. Introduce laser radar technology, use laser beams to scan the surrounding environment with high precision and build a three-dimensional environmental map; combine with visual cameras, use deep learning algorithms to analyze images in real time, and identify the shape, category and motion state of objects; by fusing multi-sensor data such as thermal, ultrasonic, laser radar, and vision, use Bayesian estimation, Kalman filtering and other algorithms to achieve more accurate environmental perception and positioning, allowing the robot to move freely in complex environments;

[0024] 3. Through the close cooperation of voice interaction submodule, gesture interaction submodule, touch screen interaction submodule, visual interaction submodule and micro-expression interaction submodule, the robot can fully adapt to different user needs and scenarios, greatly improve the service quality, shorten the distance with users, and bring users excellent intelligent interaction experience;

[0025] 4. Add an expression display screen on the top of the robot to display rich expression animations. When the robot communicates with the user, it generates corresponding expressions such as happiness, doubt, and thinking in real time based on voice intonation, semantic content and other information, allowing the user to intuitively feel the robot's "emotional" changes, enhancing the fun and intimacy of human-computer interaction;

[0026] 5. Add a placement frame to the robot so that the robot can carry and move items during movement, improving the convenience of use for users. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 This is a three-dimensional diagram of the overall structure of a pan-tech information service robot based on AI artificial intelligence proposed by the present invention;

[0028] Figure 2 This is a three-dimensional diagram of the overall structure of a pan-tech information service robot based on AI artificial intelligence proposed by the present invention;

[0029] Figure 3 This is a three-dimensional diagram of the overall structure of a pan-tech information service robot based on AI artificial intelligence proposed by the present invention;

[0030] Figure 4 A three-dimensional diagram of the structure of a visualization component of a pan-tech information service robot based on AI artificial intelligence proposed by the present invention;

[0031] Figure 5 This is a logic block diagram of the internal system of a pan-tech information service robot based on AI artificial intelligence proposed by the present invention.

[0032] In the figure: 1. Robot body; 2. Visualization component; 201. Vertical board; 202. Support board; 203. Display host; 204. Expression display screen; 205. Touch screen; 3. Placement component; 301. Vertical board; 302. Inclined board; 303. Limit rod; 304. Placement frame; 4. Multi-sensor fusion component; 401. LiDAR; 402. Visual camera; 403. Thermal sensor; 404. Ultrasonic sensor; 5. Mobile component; 501. Base; 502. Universal wheel; 503. Drive motor; 504. Placement block; 6. Speaker; 7. Microphone. DETAILED DESCRIPTION

[0033] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments.

[0034] Embodiment 1:

[0035] Reference Figure 1-5 A pan-tech information service robot based on AI artificial intelligence includes a robot body 1, the robot body 1 includes hardware equipment and a software system, the hardware equipment includes a visualization component 2, a placement component 3, a multi-sensor fusion component 4 and a movement component 5, and the software system includes a multi-sensor fusion module, a human-computer interaction module, a visualization information display module and a processing module.

[0036] The multi-sensor fusion module is used to enable the robot body 1 to accurately perceive the overall environment and empower the robot body 1 to make intelligent decisions. The data collected by the multi-sensor fusion component 4 is processed and then displayed through the visual information display module. The multi-sensor fusion module includes a data acquisition sub-module and a data fusion processing sub-module. The data acquisition sub-module includes a lidar acquisition unit, a visual camera acquisition unit, a thermal sensor acquisition unit and an ultrasonic sensor acquisition unit.

[0037] The laser radar acquisition unit is based on the laser emission and reception control algorithm and the point cloud data construction algorithm. The visual camera acquisition unit is based on the autofocus and exposure control algorithm and the data compression algorithm. The thermal sensor acquisition unit is based on the temperature calculation and compensation algorithm, the filtering and denoising algorithm, the image reconstruction and enhancement algorithm, and the target recognition and tracking algorithm. The ultrasonic sensor acquisition unit is based on the transit time algorithm and the multiple measurement averaging algorithm.

[0038] The lidar acquisition unit, visual camera acquisition unit, thermal sensor acquisition unit and ultrasonic sensor acquisition unit in the multi-sensor fusion module start data acquisition simultaneously.

[0039] In this embodiment, it should be noted that: the multi-sensor fusion module plays a key role in the robot's environmental perception and intelligent decision-making; the human-computer interaction module improves the customer's usage experience and meets diverse usage needs; the visual information display module displays abstract knowledge in a concrete way, and can perform abnormal warnings and troubleshooting, and can also enhance the experience of the human-computer interaction module; the processing module plays a role in coordination and calculation in the operation of the entire software system.

[0040] The LiDAR 401 emits laser beams in all directions at a frequency of up to 100,000 times per second. Based on the time-of-flight (ToF) principle, it accurately measures the time difference between the round trip of light, thereby obtaining the distance information of surrounding objects and constructing a high-precision three-dimensional point cloud map with an accuracy of up to millimeter level, which can meticulously present the contours and structural layout of objects in the environment.

[0041] The factory environment is complex. The robot body 1 uses laser radar to build a map, which can accurately plan the optimal path from the starting point to the target point, avoid obstacles, and efficiently complete the movement and transportation tasks.

[0042] The visual camera 402 uses a high-resolution CMOS image sensor with excellent sensitivity and low noise characteristics, and can capture color images of the surrounding environment in real time. After acquisition, it immediately uses pre-processing algorithms such as grayscale, Gaussian filtering, and image enhancement to remove noise interference in the image, improve image clarity, and provide high-quality materials for subsequent target recognition.

[0043] It can still capture clear images in low-light environments (such as factories with low lighting or factories with many items causing dim lighting), effectively suppress noise interference, ensure image quality, process factory environment pictures in real time, and cooperate better with the visual interaction sub-module and micro-expression interaction sub-module in the human-computer interaction module to improve operator use.

[0044] Thermal sensor 403 is highly sensitive to environmental heat sources by virtue of special quantum tunneling effect thermosensitive materials, continuously monitors temperature distribution, and outputs accurate temperature data. Its optical filter and intelligent gain control technology effectively deal with strong light interference, ensuring that heat sources can still be accurately detected in complex lighting environments (such as industrial workshops with strong light), providing key support for equipment failure warning and environmental temperature monitoring.

[0045] The ultrasonic sensor 404 adaptively adjusts the sound wave emission frequency according to the environmental conditions (such as dust concentration and humidity), emits sound waves and judges the distance of surrounding obstacles based on the echo. In dusty construction sites, the frequency is intelligently reduced to enhance the penetration of sound waves, ensure the reliability of the obstacle avoidance function, and provide timely and accurate information for the robot body 1 to avoid obstacles at close range.

[0046] After the data acquisition submodule collects the data from the above sensors, it uses the Kalman filter algorithm to denoise and optimize the state estimation of each sensor data.

[0047] Taking mobile target tracking as an example, by combining the distance information of the laser radar 401 and the image features of the visual camera 402, the Kalman filter continuously predicts the target's position, speed and other state variables, filters out measurement noise, and makes the target trajectory smoother and more accurate.

[0048] Next, the probability fusion model based on Bayesian estimation plays a key role. When judging whether a certain area is potentially dangerous (such as judging whether the heat source is caused by equipment failure), the temperature probability distribution provided by the thermal sensor 403, the probability of judging the abnormal appearance of the equipment by the visual camera 402, and the auxiliary information of other sensors are fused through the Bayesian theorem to obtain a more accurate posterior probability, thereby improving the level of intelligent decision-making in environmental perception.

[0049] At the same time, the simultaneous localization and mapping (SLAM) algorithm is used to deeply integrate the point cloud data of the lidar 401 and the image features of the visual camera 402. During the movement of the robot body 1, its own position estimation is continuously optimized, and the environmental map is updated in real time to achieve high-precision autonomous positioning and environmental cognition. The error can be controlled within the centimeter level, ensuring that the robot body can move unimpeded in complex dynamic environments.

[0050] Embodiment 2:

[0051] Different from the first embodiment, Figure 1-5 , this embodiment also has the following further content: the human-computer interaction module is used to improve the customer's usage needs, and the human-computer interaction module includes a voice interaction sub-module, a gesture interaction sub-module, a touch screen interaction sub-module, a visual interaction sub-module and a micro-expression interaction sub-module.

[0052] The visual information display module is used to present abstract knowledge in a concrete way, and can also provide abnormal warning and troubleshooting. The visual information display module can enhance the experience of the human-computer interaction module.

[0053] The voice interaction submodule is based on voice recognition algorithm and speech synthesis algorithm, the gesture interaction submodule is based on gesture recognition algorithm and gesture interaction feedback algorithm, the touch screen interaction submodule is based on touch operation response algorithm and task progress display algorithm, the visual interaction submodule is based on expression animation generation algorithm and knowledge visualization display algorithm, and the micro-expression interaction submodule is based on micro-expression capture algorithm, micro-expression recognition algorithm and micro-expression interaction application algorithm.

[0054] In this embodiment, it should be noted that the voice interaction submodule uses a deep learning framework, such as TensorFlow or PyTorch, to build a model architecture that combines a convolutional neural network (CNN) with a long short-term memory network (LSTM).

[0055] CNN is responsible for extracting the spectral features of speech and accurately capturing the frequency and amplitude changes of speech; LSTM is good at processing the timing information of speech and remembering the logical relationship between speech. The two work together to greatly improve the model's ability to understand speech. When users issue instructions in natural language, such as "help me get XX items" and "transport the items to the XX area of ​​the factory", the speech recognition module quickly converts the speech signal into text with an accuracy rate of more than 95%, and can accurately recognize even in noisy public places (such as the environment where the noise of people in the factory, the noise of machine startup, and the sound of object handling are intertwined).

[0056] The gesture recognition algorithm first preprocesses the hand depth image collected by the visual camera 402, including image smoothing, denoising and other operations to improve the image quality. Then, a variety of feature extraction methods based on contours, key points, etc. are used. For example, the hand area is separated from the background through the depth threshold segmentation algorithm, the hand contour features are extracted, and the approximate position and shape of the fingers are determined; at the same time, the image is feature extracted using a convolutional neural network (CNN). CNN can automatically learn the deep features of the hand under different gesture postures, such as the degree of finger bending, palm direction, etc. These features are converted into high-dimensional vectors to provide a basis for subsequent gesture classification. A variety of classification algorithms such as support vector machines (SVM) and deep learning classifiers are used to classify and identify gesture feature vectors. In the training stage, a large amount of sample data containing different gesture actions is collected to train the classifier. For example, in a factory environment, the environment is noisy and the user speaks in a low voice, and instructions can be issued to the robot body 1 through gestures.

[0057] The touch screen interaction submodule follows the principles of ergonomics and cognitive psychology, adopts high contrast and light-adaptive color matching, arranges the core functional areas according to the "golden ratio", and guides the operation with dynamic effects. It integrates intelligent prediction functions to preload information based on user historical habits to improve the smoothness of operation. It uses augmented reality (AR) and mixed reality (MR) technologies combined with 5G low-latency networks to achieve immersive task progress visualization.

[0058] The visual interaction submodule uses real-time motion capture and artificial intelligence-driven animation technology to collect the robot's own "posture" through micro motion capture sensors on key parts of the robot's body, and uses deep neural network algorithms to generate expression animations that fit the situation by combining voice, semantics, and scene information. The cultural difference adaptation mechanism is introduced to support emotional response in group interaction scenes, so that the robot body 1 can show appropriate expressions through the expression display screen 204 in different communication atmospheres, enhancing the realism of social interaction.

[0059] The micro-expression interaction submodule uses the high-definition visual camera 402 array on the front of the robot body 1, combined with intelligent anti-shake and image stitching algorithms, to lock the user's face from multiple angles, with a resolution of 1080p or above and a frame rate of 60 frames per second. It is assisted by near-infrared light imaging that penetrates the shallow layer of the skin to accurately capture micro-expression details such as the corners of the mouth rising and the corners of the eyes shrinking. Based on the Facial Action Coding System (FACS) fusion deep learning model, such as the Transformer-based multi-head attention network, micro-expression patterns are learned; the dynamic time warping (DTW) algorithm is used to overcome individual differences, flexibly match time series, and accurately judge the emotional meaning behind micro-expressions.

[0060] Embodiment three:

[0061] The difference from the first and second embodiments is that, referring to Figure 1-4 , this embodiment also has the following further contents: the visualization component 2 is used to improve the interactivity of the user, the visualization component 2 includes a stand 201, a support plate 202, a display host 203, an expression display screen 204 and a touch screen 205, a stand 201 is provided on one side of the robot body 1, a support plate 202 is provided on the top of the stand 201, a display host 203 is provided on the top of the support plate 202, a touch screen 205 is provided on the side of the stand 201 away from the robot body 1, and an expression display screen 204 is provided on the side of the display host 203 close to the touch screen 205, the multi-sensor fusion component 4 The visualization component 2 is used to improve the user's interactivity and enable the robot body 1 to better judge the environment, avoid risks and issue early warnings. The multi-sensor fusion component 4 includes a laser radar 401, a visual camera 402, a thermal sensor 403 and an ultrasonic sensor 404. The laser radar 401 is provided on the top of the display host 203, the visual camera 402 is provided on the top of the touch display screen 205, the thermal sensors 403 are provided on the four sides of the placement block 504 and on both sides of the display host 203, and the laser radar 401 is provided around the bottom of the base 501 near the universal wheel 502.

[0062] Speakers 6 are provided inside the vertical plate 301 and the inclined plate 302, a microphone 7 is provided on the side of the touch display screen 205, the base 501 and the placement block 504 are arranged in a rectangular shape, and the four corners of the rectangle of the base 501 and the placement block 504 are arranged in an arc shape.

[0063] In this embodiment, it should be noted that the laser radar 401 is placed on the top of the robot body 1. From the perspective of vision, this position can provide a 360-degree all-round horizontal scanning range without blind spots, and can capture the distribution of all objects and personnel around, providing the most comprehensive data support for accurate navigation and path planning. For example, in a factory scene, the robot body needs to be aware of workers and objects moving in all directions at any time. The laser radar 401 at the top center can ensure that it can quickly detect and make corresponding guiding actions to avoid collisions.

[0064] This position is relatively high, which reduces the interference of the robot's own structure on the propagation of the laser beam, ensures the integrity of the point cloud data obtained by the laser radar 401, and improves the accuracy of environmental perception.

[0065] When the robot body 1 moves via the moving component 5, the laser radar 401, the visual camera 402, the thermal sensor 403 and the ultrasonic sensor 404 scan and read the surrounding environment in real time. The user issues instructions to the robot body 1 by clicking the operation button on the touch display 205. When the thermal sensor 403 detects abnormal temperature, an alarm is sounded through the speaker 6. At the same time, the robot body 1 can read the data through the speaker 6 to improve the user's comfort.

[0066] The user's behavior is scanned through the microphone 7 and the visual camera 402 on the top of the touch display screen 205, and the voice is collected through the microphone 7 and data is collected and analyzed.

[0067] In a noisy factory environment or when the user's voice is low, the robot body 1 can be operated through fixed gestures. The gestures and the user's micro-expressions are scanned by the visual camera 402 and combined with intelligent anti-shake and image stitching algorithms for data analysis to improve the human-computer interaction performance.

[0068] Embodiment 4:

[0069] Reference Figure 1-4Compared with the first, second and third embodiments, in the present embodiment: the moving component 5 is used for the robot body 1 to move and rotate at multiple angles during use, the moving component 5 includes a base 501, universal wheels 502, a drive motor 503 and a placement block 504, the bottom end of the robot body 1 is provided with a base 501, the bottom end of the base 501 and the four corners are provided with universal wheels 502, the middle of the top end of the base 501 is provided with a drive motor 503, and the execution end of the drive motor 503 is connected to the placement block 504.

[0070] The placement component 3 is used for the user to place some items on the robot body 1 for transportation during use. The placement component 3 includes a vertical plate 301, an inclined plate 302, a limit rod 303 and a placement frame 304. The vertical plate 301 is provided on the side of the vertical plate 201 away from the touch display screen 205 and at the top of the placement block 504. The bottom end of the vertical plate 301 is connected to the inclined plate 302. The top of the placement block 504 away from the inclined plate 302 is provided with a limit rod 303. The limit rod 303 is provided with multiple placement frames 304 in sequence from top to bottom.

[0071] The support plate 202 is arranged at the top of the vertical plate 301 and the inclined plate 302 . The support plate 202 , the vertical plate 301 and the inclined plate 302 are arranged in a triangular shape. The side of the placement frame 304 away from the limiting rod 303 is fixedly connected to the inclined plate 302 .

[0072] In this embodiment, it should be noted that: when the robot body 1 is used in a factory, when user A needs to transport parts, documents or items, he can place them in the placement frame 304 on the rear side of the robot body 1, issue instructions, and the robot body 1 will transport them.

Claims

1. A pan-tech information service robot based on AI artificial intelligence, comprising a robot body (1), characterized in that: The robot body (1) includes hardware equipment and a software system, wherein the hardware equipment includes a visualization component (2), a placement component (3), a multi-sensor fusion component (4) and a movement component (5); The software system includes a multi-sensor fusion module, a human-computer interaction module, a visual information display module and a processing module; The multi-sensor fusion module is used to enable the robot body (1) to accurately perceive the overall environment and enable the robot body (1) to make intelligent decisions. The data collected by the multi-sensor fusion component (4) is processed and then displayed through the visual information display module. The multi-sensor fusion module includes a data collection submodule and a data fusion processing submodule. The data collection submodule includes a laser radar collection unit, a visual camera collection unit, a thermal sensor collection unit, and an ultrasonic sensor collection unit. The human-computer interaction module is used to improve the usage needs of customers. The human-computer interaction module includes a voice interaction submodule, a gesture interaction submodule, a touch screen interaction submodule, a visual interaction submodule and a micro-expression interaction submodule; The visual information display module is used to present abstract knowledge in a concrete way, and can also perform abnormal warning and troubleshooting. The visual information display module can enhance the experience of the human-computer interaction module.

2. The AI-based pan-tech information service robot according to claim 1, characterized in that: The visualization component (2) is used to improve the interactivity of the user, and the visualization component (2) comprises a vertical plate (201), a support plate (202), a display host (203), an expression display screen (204) and a touch screen (205); the vertical plate (201) is provided on one side of the robot body (1); the support plate (202) is provided on the top of the vertical plate (201); the display host (203) is provided on the top of the support plate (202); the touch screen (205) is provided on the side of the vertical plate (201) away from the robot body (1); and the expression display screen (204) is provided on the side of the display host (203) close to the touch screen (205); The moving assembly (5) is used for the robot body (1) to move and rotate at multiple angles during use. The moving assembly (5) comprises a base (501), universal wheels (502), a driving motor (503) and a placement block (504). The bottom end of the robot body (1) is provided with a base (501), the bottom end of the base (501) and the four corners thereof are provided with universal wheels (502), the middle of the top end of the base (501) is provided with a driving motor (503), and the execution end of the driving motor (503) is connected to the placement block (504); The placement component (3) is used for a user to place some articles on the robot body (1) for transportation during use. The placement component (3) comprises a vertical plate (301), an inclined plate (302), a limiting rod (303) and a placement frame (304). The vertical plate (301) is provided on a side of the vertical plate (201) away from the touch display screen (205) and at the top of the placement block (504). The bottom end of the vertical plate (301) is connected to the inclined plate (302). The limiting rod (303) is provided on the top of a side of the placement block (504) away from the inclined plate (302). The limiting rod (303) is provided with a plurality of placement frames (304) in sequence from top to bottom. The multi-sensor fusion component (4) cooperates with the visualization component (2) to improve the interactivity of the user, and at the same time enables the robot body (1) to better judge the environment, avoid risks and issue warnings. The multi-sensor fusion component (4) includes a laser radar (401), a visual camera (402), a thermal sensor (403) and an ultrasonic sensor (404). The top of the display host (203) is provided with a laser radar (401), the top of the touch display screen (205) is provided with a visual camera (402), the four surfaces of the placement block (504) and the two sides of the display host (203) are provided with thermal sensors (403), and the bottom of the base (501) is provided with a laser radar (401) near the position of the universal wheel (502).

3. The AI-based pan-tech information service robot according to claim 1, characterized in that: The voice interaction submodule is based on a voice recognition algorithm and a voice synthesis algorithm, the gesture interaction submodule is based on a gesture recognition algorithm and a gesture interaction feedback algorithm, the touch screen interaction submodule is based on a touch operation response algorithm and a task progress display algorithm, the visual interaction submodule is based on an expression animation generation algorithm and a knowledge visualization display algorithm, and the micro-expression interaction submodule is based on a micro-expression capture algorithm, a micro-expression recognition algorithm and a micro-expression interaction application algorithm.

4. The AI-based pan-tech information service robot according to claim 3, characterized in that: The laser radar acquisition unit is based on a laser emission and reception control algorithm and a point cloud data construction algorithm; the visual camera acquisition unit is based on an automatic focus and exposure control algorithm and a data compression algorithm; the thermal sensor acquisition unit is based on a temperature calculation and compensation algorithm, a filtering and denoising algorithm, an image reconstruction and enhancement algorithm, and a target recognition and tracking algorithm; and the ultrasonic sensor acquisition unit is based on a transit time algorithm and an algorithm for taking an average value of multiple measurements.

5. The AI-based pan-tech information service robot according to claim 4, characterized in that: The laser radar acquisition unit, visual camera acquisition unit, thermal sensor acquisition unit and ultrasonic sensor acquisition unit in the multi-sensor fusion module synchronously start data acquisition work.

6. The AI-based pan-tech information service robot according to claim 2, characterized in that: The support plate (202) is arranged at the top ends of the vertical plate (301) and the inclined plate (302); the support plate (202), the vertical plate (301) and the inclined plate (302) are arranged in a triangular shape.

7. The AI-based pan-tech information service robot according to claim 6, characterized in that: A speaker (6) is provided inside the vertical plate (301) and the inclined plate (302), and a microphone (7) is provided on the side of the touch display screen (205).

8. The AI-based pan-tech information service robot according to claim 7, characterized in that: The side of the placement frame (304) away from the limiting rod (303) is fixedly connected to the inclined plate (302).

9. The AI-based pan-tech information service robot according to claim 7, characterized in that: The base (501) and the placement block (504) are arranged in a rectangular shape, and the four corners of the base (501) and the placement block (504) are arranged in an arc shape.

Images