Intelligent teaching desk system based on multi-mode unmanned aerial vehicle

Abstract

The invention discloses an intelligent teaching desk system based on a multi-mode unmanned aerial vehicle, and belongs to the technical field of education equipment. The system comprises a teacher's desk main body integrated with a hidden unmanned aerial vehicle cabin, a folding unmanned aerial vehicle equipped with a multispectral sensor, and an edge calculation center, and innovatively realizes three functions: 1, intelligent patrol: the unmanned aerial vehicle performs autonomous navigation through laser SLAM (mapping precision + / -3cm); millimeter wave radar vital sign detection (respiratory frequency precision + / -0.5 times / min) and thermal imaging anomaly positioning are fused to replace manual patrol. 2, teaching enhancement: supporting AR knowledge point projection (the brightness is 500nit) and a multi-mode question and answer system (the dialect recognition rate is greater than or equal to 95%), and realizing air touch interaction through an ultrasonic array; and third, five-level emergency response: from sign abnormity early warning to autonomous escape way opening, a spraying system and AED equipment guidance are linked, and the response delay is less than or equal to 2.5 seconds.

Description

Technical Field

[0001] The present invention belongs to the technical field of educational equipment, and particularly relates to an intelligent teaching desk system based on a multi-modal drone, which realizes the integrated functions of classroom dynamic monitoring, enhanced teacher-student interaction, and emergency response. Background Art

[0002] Existing teaching equipment has the following technical defects: traditional desks have a single function and lack dynamic interaction capabilities; classroom inspections rely on manual or fixed cameras, resulting in monitoring blind spots; the efficiency of emergency response is low, and there is a lack of proactive handling means; the integration of intelligent devices in teaching scenarios is insufficient. Summary of the Invention

[0003] The composition of an intelligent teaching desk system based on a multi-modal drone (attached Figure 1 ) is as follows, which includes three parts: an intelligent desk main body, a multi-modal drone, and a control center.

[0004] The intelligent desk main body is composed of a steel-wood composite structure box (size 1200×600×900mm), a hidden drone cabin (equipped with a temperature control and automatic charging module), and an electric slide rail cabin cover (opening and closing angle ≥ 120°, noise ≤ 40dB).

[0005] The key components and functional parameters of the multi-modal drone are: a four-axis folding fuselage (expanded diameter 380mm), a three-axis stabilized gimbal (integrated with a 1080P camera + infrared thermal imaging), a directional microphone array (pickup radius 8m), and a lidar obstacle avoidance system (detection distance 0.2 - 5m).

[0006] The key components and functional parameters of the control center are: an edge computing unit (NPU computing power 4TOPS), a multi-device communication gateway (supporting Wi-Fi6 + Zigbee3.0), and an emergency response management module.

[0007] The core innovation points of the intelligent teaching desk system based on a multi-modal drone are reflected in the space adaptive storage system, intelligent teaching assistance function, and five-level emergency response mechanism.

[0008] Space adaptive storage system: Adopting the concept of a folding mechanism, a three-stage telescopic landing gear is designed; the layout inside the cabin adopts a blast diversion structure to achieve contactless takeoff and landing; an electric flipping mechanism is integrated, and the opening and closing speed of the cabin cover is adjustable.

[0009] Intelligent teaching assistance function: Classroom concentration analysis: Integrating visual gesture recognition and voiceprint emotion detection; dynamic knowledge point annotation: The drone projects AR teaching content to a specific area; multi-modal Q&A system: Realizing natural interaction between teachers and students through voice + gesture.

[0010] The five - level emergency response mechanism includes levels 1 to 5. Level 1: Early warning of abnormal student behavior (triggered when the leaving - seat rate > 30%); Level 2: Detection of sudden illness (body temperature > 38°C + posture fall determination); Level 3: Fire recognition (thermal imaging > 60°C + smoke sensor linkage); Level 4: Emergency evidence collection (automatically tracking abnormal targets and recording videos); Level 5: Autonomous escape guidance (acoustic - optical indication + door - window control linkage). Brief Description of the Drawings

[0011] Figure 1 : Schematic diagram of the system structure. Component descriptions in the figure are as follows: Dual - light camera: combination of visible light and thermal imaging, supporting 5 - fold optical zoom; Millimeter - wave radar: 60GHz frequency band, detection range 0.2 - 5m, breathing detection accuracy ±0.5 times / minute; Communication gateway: supports Zigbee3.0 / Wi - Fi6 / 5G private network slicing technology.

[0012] Figure 2 : Flowchart of multi - modal data fusion. Technical characteristics of the picture are as follows: Space - time alignment accuracy: 10ms - level synchronization; Fusion algorithm: dynamic weight adjustment mechanism, supporting complementary verification of radar / visual data; Cheating recognition: pose detection based on YOLOv5 + abnormal voiceprint analysis.

[0013] Figure 3 : State transition diagram of emergency response. Related state descriptions in the figure are as follows: Device linkage: including door - window control, security alarm, and first - aid material delivery; Manual intervention: displaying a 3D situation map through the teacher - side AR interface.

[0014] Figure 4 : Schematic diagram of the cheating recognition interface in the examination room mode. Interface element descriptions in the figure are as follows: Heat map: representing the density of suspicious behaviors with color gradients; Pose recognition: detecting abnormal limb movements (turning head frequency > 2 times / minute); Alarm panel: displaying candidate ID, abnormal type, and confidence level. Detailed Implementation Modes

[0015] Example 1: Regular teaching scenario. First step: The teacher clicks the control panel on the lectern to start the "classroom inspection mode"; Second step: The hatch automatically opens, and the drone takes off vertically to hover at 2.5m; Third step: Through the AI algorithm, it completes: student attendance (face recognition accuracy 99.8%), concentration analysis (combining eye tracking and pose recognition), and generation of heat maps for key and difficult knowledge points.

[0016] Example 2: Emergency in a chemistry experiment. First, the thermal imaging detects abnormal temperature ( > 80°C) on a certain experimental bench; Then the drone autonomously flies to the target area: starts high - definition zoom to confirm the hazard source, broadcasts a voice warning (volume automatically increases to 90dB), and guides the location of the nearest fire extinguisher (AR projection indication); Finally, synchronously uploads the accident data to the campus security platform.

Claims

1. An intelligent teaching desk system based on a multi-modal drone, characterized in that Includes: a lecture table body with an integrated hidden drone cabin; a foldable drone with multi-modal perception capabilities; a central control module that supports edge computing; and a multi-protocol communication interface (HDMI / USB-C / RS485).

2. The system as claimed in claim 1, wherein the lecture table structure comprises: a) Cabin environment control system (temperature 20-25℃ / humidity ≤60%RH); b) Electromagnetic adsorption charging interface (charging efficiency ≥ 90%); c) Anti-accidental touch safety lock mechanism (pressure sensing + dual authentication).

3. The system as claimed in claim 1, wherein the drone module comprises: a) a detachable teaching tool cabin (accommodating a laser pen / loudspeaker, etc.); b) a dynamic zoom camera (5x optical zoom); c) an ultrasonic cleaning device (automatic dust removal cycle ≤ 24h). The system as claimed in claim 1, wherein the control method thereof comprises: a) automatic recognition of teaching scene modes (theoretical class / laboratory class / exam); b) student position heat map generation algorithm; c) multi-device audio synchronization delay compensation (≤50ms).

4. The system as described in claim 1, wherein the emergency protocol includes: a) automatically opening doors and windows and activating the sprinkler system in case of fire; b) triggering AED device location guidance in case of sudden illness; c) local encrypted storage (AES-256) of emergency video data.

5. The system according to claim 1, characterized in that Supports: a) AR content linkage projection with electronic whiteboard; b) cloud synchronization of teaching data (compliant with LTI 1.3 standard); c) multi-classroom equipment network management (supports up to 32 nodes).

6. The system as claimed in claim 1, wherein the interaction method comprises: a) Gesture control (supports 12 standard actions in ISO gesture library) 'b) Voice wake-up (dialect recognition accuracy ≥ 95%); c) Linkage with teachers’ wearable devices (smart watches / AR glasses).

7. The system as described in claim 1, its hardware configuration includes: a) the main control of the lecture desk adopts Rockchip RK3588S chip; the drone is equipped with Qualcomm QCS6490 processor; the communication module supports 5G private network slicing technology.

8. The system according to claim 1, characterized in that Extended functions: a) Start cheating behavior recognition algorithm in examination room mode; b) Real-time warning of dangerous operations in laboratory classes; c) Teaching environment quality monitoring (PM2.5 / CO2 / illuminance).

Images