Multi-mode unmanned aerial vehicle cluster nursing system and deployment platform thereof

Abstract

The invention discloses a multi-scene unmanned aerial vehicle nursing system, and belongs to the technical field of intelligent health care. The system comprises a modular guarding work station, a multi-mode unmanned aerial vehicle cluster and a central scheduling platform, and innovatively realizes three core functions: first, elastic deployment: dynamically adjusting the unmanned aerial vehicle ratio (1: 4-1: 8) based on the number of beds, adopting a tidal charging strategy to guarantee the endurance of a high-risk area, and covering eight beds by a single machine, so that the equipment cost is greatly reduced; secondly, non-contact monitoring: fusing 60GHz millimeter wave radar and a thermal imaging technology to realize monitoring of vital signs outside 2 meters, and identifying 9 types of abnormal sounds such as pain and moan through voiceprint analysis; and thirdly, a fifth-level emergency system: the response speed is obviously improved compared with a traditional scheme from automatic recheck to autonomous emergency channel opening. The system is suitable for nursing homes, hospitals, kindergartens, pet fostering centers and other places needing continuous monitoring, and organic integration of intelligent patrol, health early warning and emergency response is achieved.

Description

Technical Field

[0001] The present invention belongs to the technical field of intelligent health care, and specifically relates to a multi-modal UAV cluster nursing system and its deployment platform, which is applicable to places that require continuous monitoring such as nursing homes, hospitals, kindergartens, pet foster care centers, etc., and realizes the organic integration of intelligent patrol, health warning and emergency response. Background Art

[0002] The existing nursing systems have the following technical defects: manual patrol has time blind spots and subjective judgment errors; fixed vital sign monitoring equipment restricts the freedom of movement of the elderly; the response to emergencies depends on personnel on-site confirmation; the deployment cost of multi-bed monitoring equipment is high. Summary of the Invention

[0003] The system of the multi-modal UAV cluster nursing system and its deployment platform (attached Figure 1 ) consists of an intelligent duty workstation, multi-modal nursing UAVs, and a central dispatcher.

[0004] Composition and functional parameters of the intelligent duty workstation: modular cabinet (supporting parallel deployment of 4 - 12 UAVs); three-dimensional charging matrix (charging efficiency ≥ 95%); environmental pre-inspection cabin (automatically disinfecting / self-checking UAVs).

[0005] Composition and functional parameters of the multi-modal nursing UAV: six-axis safety protection structure (diameter 450mm ± 5%); expandable function cabin (medicine / emergency supplies / disinfection module); multi-spectral sensing array: 60GHz millimeter-wave radar (breathing / heart rate monitoring), dual-light gimbal (visible light + thermal imaging), ToF fall detection sensor (accuracy ± 2cm), gas monitoring module (CO 2 / methane / VOC).

[0006] Composition and functional parameters of the central dispatching system: dynamic path planning engine (optimized by RRT* algorithm); digital twin platform for bed status; multi-level early warning management protocol (5-level response mechanism).

[0007] The core innovation points of the present invention are reflected in three aspects: elastic deployment mechanism, non-contact monitoring, and five-in-one emergency system.

[0008] Elastic deployment mechanism: bed number and UAV ratio algorithm (dynamically adjusted from 1:4 to 1:8); tidal charging strategy (prioritizing to ensure the endurance of high-risk areas); cross-floor relay cruise (realizing signal coverage through relay base stations).

[0009] Non-contact monitoring: integrating radar monitoring technology to realize detection of vital signs more than 2 meters away; detecting sudden symptoms such as epilepsy based on micro-motion recognition algorithm; using voiceprint analysis to identify abnormal sounds such as painful groans.

[0010] The five-in-one emergency system includes levels 1 to 5. Level 1: Automatic re-examination of abnormal signs (error rate < 0.5%); Level 2: Activate adjacent drones for collaborative observation; Level 3: Dispense emergency drugs (positioning accuracy ±15 cm); Level 4: Link the security system (access control / lighting / calling); Level 5: Independently open an emergency passage (dynamic planning of evacuation routes).

[0011] Analysis of the technical advantages of the present invention. First, in terms of cost control, a single drone covers 8 beds, and the equipment investment is reduced by 62%; adopting a tidal charging strategy, the energy consumption is reduced by 35%; reusing the existing building Wi-Fi network without the need for separate wiring. Secondly, it is reflected in enhanced safety: integrating a triple verification mechanism, the false alarm rate < 0.1%, adopting an emergency passage algorithm, the response speed is increased by 40%, and the mechanical protection meets the ISO 13482 safety standard.

[0012] The present invention can also be extended in the following scenarios: adapting to scenarios such as hospitals / kindergartens by replacing the functional cabin, the software platform supports one-key switching of multi-site modes, and the data interface is compatible with mainstream medical / security systems. Description of the Drawings

[0013] Figure 1 : Schematic diagram of the system structure. The description of the core components in the figure is as follows. Millimeter-wave radar: TI IWR6843 chip, respiratory detection accuracy ±0.3 times / minute; Dual-light pan-tilt: Visible light 1080P@60fps + Thermal imaging 256×192 resolution; Charging matrix: XYZ three-axis robotic arm positioning, charging efficiency ≥95%.

[0014] Figure 2 : Flowchart of multi-modal data fusion. The relevant technical characteristics involved in the figure are as follows. Space-time alignment: UWB positioning (accuracy ±10 cm) + Hardware clock synchronization; Fusion weight: Dynamically adjusted (radar data weight 0.6 / vision 0.3 / environment 0.1); Response delay: From detection to instruction issuance ≤800 ms.

[0015] Figure 3 : State transition diagram of emergency response. The response levels in the figure are as follows. Level 1: Local audible and visual warning; Level 3: Drone approaches for reconnaissance; Level 5: Link the fire / medical system.

[0016] Figure 4 : Flowchart of the night patrol work in the nursing home. The key parameters of the relevant processes in the figure are as follows. Scanning interval: Stay at each bed ≥30 seconds; Accuracy of antipyretic patch dispensing: ±15 cm; Review response time: ≤20 seconds.

[0017] Figure 5: Flowchart of kindergarten safety guardianship work. Relevant technical indicators of the engineering process in the figure: Voiceprint recognition response: ≤1.2 seconds; Music volume fade: 65dB → 50dB (attenuation in 10 seconds); Warning circle radius: 3m (error ±0.5m). Specific implementation manner

[0018] Example 1: Night patrol in a nursing home. The drone starts full-floor scanning according to the preset route; the millimeter-wave radar penetrates the curtain to detect vital signs; it is found that the breathing rate in Room 305 is abnormal (28 times / min): Automatically dispatch the No. 2 drone for multi-angle verification, and the thermal imaging confirms that the body temperature has risen (38.2°C), place a fever-reducing patch on the bedside table (error ±10cm), wake up the duty nurse station and push the patient's file.

[0019] Example 2: Kindergarten safety guardianship. Voiceprint analysis to identify abnormal crying sounds; The drone swarm starts the following protocols: a) The host locks the sound source position and records evidence, b) The auxiliary drones form an isolation warning circle, c) Automatically play soothing children's songs (volume fade adjustment); Synchronously push the event report to the principal's terminal.

Claims

1. A multi-modal drone swarm care system and its deployment platform system, characterized in that Includes: modular drone duty workstation, drone cluster with vital signs monitoring function, central dispatching platform supporting digital twins, and cross-system linkage interface (HL7 / Modbus).

2. The system as claimed in claim 1, wherein the workstation structure comprises: a) Three-dimensional charging matrix (XYZ three-axis robotic arm positioning); b) UV + ozone dual disinfection module; c) Rapid loading channel for emergency supplies.

3. The system as claimed in claim 1, wherein the drone module comprises: a) a quick-detachable functional expansion cabin (compatible with medicine / firefighting / disinfection modules); b) a laser SLAM autonomous navigation system (mapping accuracy ±3cm); c) an automatic ejection device for a safety net.

4. The system as claimed in claim 1, wherein the scheduling method comprises: a) Zoning cruise algorithm based on care level (VIP / ordinary / isolation); b) Multi-aircraft task allocation optimization model (minimizing total flight distance); c) Low battery priority return strategy (triggered when 20% battery remaining).

5. The system as claimed in claim 1, wherein the monitoring protocol includes: a) abnormal respiratory rate detection (<8 times / min or >30 times / min); b) bed leaving timeout warning (>30 minutes without detecting vital signs); c) excrement retention identification (image analysis + odor sensing).

6. The system according to claim 1, characterized in that it supports: a) Cross-validation with the vital signs data of the medical mattress; b) Pushing structured nursing records to the HIS system; c) Pet abnormal behavior library (scratching / barking / hunger strike).

7. The system as claimed in claim 1, wherein the interaction method comprises: a) Voice soothing system (supports dialects / children's intonation); b) Emergency call red light positioning (brightness adjustable 300-1000 lux); c) Medication reminder AR projection (time window error ±1min).

8. The system as described in claim 1, wherein the hardware configuration includes: a) using an NVIDIA Jetson Orin NX edge computing unit; b) equipped with a TI IWR6843 millimeter wave radar chip; c) integrating an AMS CCS811 gas sensor.

9. The system according to claim 1, characterized in that Extended functions: a) Early warning of abnormal gathering of children in kindergartens; b) Automatic feeding and activity monitoring of pets in foster care; c) Delivery of sterile supplies in hospital wards.

Images