CN-122025049-A - Low-altitude medical service system and method based on AI+unmanned aerial vehicle

CN122025049ACN 122025049 ACN122025049 ACN 122025049ACN-122025049-A

Abstract

The invention belongs to the technical field of intelligent medical services, and particularly relates to a low-altitude medical service system and method based on an AI+unmanned aerial vehicle. The low-altitude medical service system based on the AI+unmanned aerial vehicle comprises a medical transportation unmanned aerial vehicle, an intelligent low-altitude junction station, an intelligent mechanical arm, an intelligent logistics distribution system, an intelligent security inspection system, a multi-mode emergency response system and a low-altitude medical intelligent dispatching platform. The low-altitude medical service method based on the AI+unmanned aerial vehicle comprises the following steps of S1, constructing a multi-mode large data center, S2, developing a large model training and service platform, and S3, constructing a high-quality medical low-altitude data set. The invention can realize the rapid point-to-point distribution of medical drugs, the omnibearing dynamic monitoring of the hospital environment and the real-time early warning of security risks, and has better market application prospect.

Inventors

FU WENTAO
SUN XIYAN
DU WENHONG
Lv Hengzhi
LI YOUMING

Assignees

桂林电子科技大学
广西产研院时空信息技术研究所有限公司

Dates

Publication Date: 20260512
Application Date: 20260126

Claims (10)

1. Ai+unmanned aerial vehicle-based low-altitude medical service system, which is characterized by comprising: the medical transportation unmanned aerial vehicle is used for accurately distributing medical first-aid articles and fire-fighting emergency articles; The intelligent low-altitude junction station is used for taking off, landing and parking of the medical transport unmanned aerial vehicle, storing and taking out medical first-aid articles and fire-fighting emergency articles, and distributing and delivering tasks for the medical transport unmanned aerial vehicle; The intelligent mechanical arm is used for providing automatic power changing of the medical transportation unmanned aerial vehicle and automatic loading and unloading of medical and emergency materials, and guaranteeing safe and sterile medical material loading and unloading processes and high-efficiency and accurate automatic power changing processes of the unmanned aerial vehicle; the intelligent logistics distribution system is used for carrying out space-time prediction and path optimization according to requirements and carrying out terminal accurate distribution through multi-sensor fusion obstacle avoidance; The intelligent security inspection system is used for carrying out medical scene anomaly identification, multi-mode fusion analysis and unmanned aerial vehicle automatic security inspection planning, automatically forming inspection reports and forming closed-loop management; The multi-mode emergency response system is used for carrying out voice recognition and NLP instruction analysis in an emergency scene, automatically carrying out scene analysis on a scene picture, carrying out multi-machine collaborative operation, and automatically scheduling and matching resources to process the emergency scene; the low-altitude medical intelligent dispatching platform integrates the intelligent logistics distribution system, the intelligent security inspection system and the multi-mode emergency response system and is used for issuing tasks to the medical transportation unmanned aerial vehicle and the intelligent low-altitude junction station and realizing the functions of flight monitoring, dispatching management, route planning, facility monitoring, task analysis and safety early warning.
2. The low-altitude medical service system based on the AI+unmanned aerial vehicle, which is disclosed in claim 1, is characterized in that the intelligent low-altitude junction station integrates a logistics connection cabinet, an intelligent mechanical arm, a multi-source environment sensor and an edge computing unit, wherein the logistics connection cabinet is used for accessing medical emergency articles and fire emergency articles, the multi-source environment sensor is used for monitoring current environment data, unmanned aerial vehicle states and material information and judging whether the unmanned aerial vehicle is allowed to take off, the edge computing unit is used for issuing a take-off instruction to a medical transportation unmanned aerial vehicle after receiving a task instruction of a total system so as to support multi-machine parallel task scheduling, the intelligent mechanical arm adopts a 'vision+laser' dual-mode positioning mode, centimeter-level positioning is realized through an AI dynamic target tracking algorithm, flexible movement is realized through the intelligent mechanical arm through a multi-axis joint structure, and the tail end is provided with a flexible clamping jaw, and actions are adjusted in real time through an AI flexible force control adjustment algorithm so as to avoid damaging medicines.
3. The low-altitude medical service system based on the AI+unmanned aerial vehicle according to claim 1, wherein the low-altitude medical intelligent scheduling platform integrates an intelligent logistics distribution system, an intelligent security inspection system and a multi-mode emergency response system, supports unmanned aerial vehicle task cooperation of multi-terminal linkage of a mobile terminal and a PC terminal, supports multi-class task list collaborative operation, performs route planning through a webpage terminal, creates multi-type business documents by a cloud, remotely initiates a flight landing instruction for an aircraft, executes a flight task and returns operation information in real time.
4. The low-altitude medical service system based on the AI+ unmanned aerial vehicle according to claim 3, wherein the intelligent logistics distribution system is used for carrying out demand prediction based on a space-time large model of an ST-transform architecture, ensuring flight safety by optimizing and fusing multisource data with an AI dynamic airspace division algorithm through an A-dynamic path, guiding and stopping through an AI visual recognition algorithm and a TOF ranging sensor, adjusting the gesture through a PID control algorithm, and ensuring accurate distribution and obstacle avoidance of the tail end through an AI flight risk early warning algorithm and fusing multisource sensors.
5. The low-altitude medical service system based on the AI+unmanned aerial vehicle is characterized in that the intelligent security inspection system builds a medical scene anomaly recognition large model, the model is divided into three parts of feature extraction, feature enhancement and classification detection, a manual marking and synthesized data supplementing mode is adopted to carry out data marking, a focusing instrument, personnel, dangerous goods and environment are subjected to anomaly recognition, the intelligent security inspection system adopts visible light and infrared multi-mode fusion analysis, improves recognition accuracy, adopts a self-adaptive algorithm, transfers a model trained in a single region to other regions, greatly reduces the cost of independent training of each region, and the intelligent security inspection system arranges the unmanned aerial vehicle to carry out automatic inspection, adopts a fixed-point cruising and random spot inspection mode, cruises and covers a key position, randomly inspects a blind area, and automatically generates a report to form closed-loop management.
6. The low-altitude medical service system based on the AI+ unmanned aerial vehicle according to claim 3 is characterized in that the multi-mode emergency response system automatically extracts key information of a noisy environment and automatically associates corresponding resources to conduct AI voice recognition and NLP instruction analysis by constructing a medical emergency field knowledge graph, the multi-mode emergency response system completes image acquisition and analysis by adopting an AI emergency dispatch algorithm, automatically recognizes wounded and dangerous sources, estimates target distances, generates a scene situation graph, pushes the scene situation graph to an emergency command platform, closes a corresponding airspace and triggers an emergency dispatch strategy, automatically dispatches matched resources based on scene analysis results, and the multi-mode emergency response system conducts multi-machine collaborative operation by developing a group intelligent decision large model based on reinforcement learning training and adopting distributed control to achieve emergency response.
7. An ai+unmanned aerial vehicle-based low-altitude medical service method for the ai+unmanned aerial vehicle-based low-altitude medical service system according to any one of claims 1 to 6, comprising the steps of: S1, constructing a multi-mode big data center; S2, large model training and service platform development; and S3, constructing a high-quality medical low-altitude data set.
8. The ai+unmanned aerial vehicle-based low-altitude medical service method according to claim 7, wherein the constructing of the multi-mode big data center in S1 specifically comprises the steps of: s11, summarizing data, wherein the data comprise unmanned aerial vehicle sensor data, hospital system desensitization data, park basic data, meteorological data and equipment states; S12, preprocessing the data, wherein the preprocessing is specifically as follows: Removing a fuzzy frame in the unmanned aerial vehicle image by adopting an AI image definition evaluation algorithm, removing outliers in the point cloud by adopting an AI statistical filtering algorithm, and identifying abnormal points in the track data by utilizing an AI time sequence abnormal detection algorithm; aligning the image, the point cloud and the track data according to time stamps to ensure that the multi-source data at the same moment corresponds to the same position, automatically marking the primary characteristics of the data to be trained through an AI pre-marking algorithm, and then manually marking the data without training through a marking platform, and automatically adding a label to the data without training; S13, adopting a Ceph distributed storage system to store data in three stages, wherein SSD stores real-time data of nearly 7 days to meet high-frequency access requirements, HDD stores historical data within 1 month to meet regular query requirements, and archive data above 1 month are stored by adopting objects and are called as required; S14, establishing a data standard system, defining a data format and a field specification, periodically carrying out data quality detection, repairing or eliminating unqualified data, and ensuring the usability of the data; S15, identifying and desensitizing patient privacy information in medical data through an AI semantic analysis algorithm, storing and transmitting and encrypting material information and track data by adopting an AI dynamic encryption algorithm, realizing hierarchical access control by adopting an AI role authority model, avoiding data unauthorized access, recording all data access operations, and supporting abnormal access behavior tracing.
9. The ai+ unmanned aerial vehicle-based low-altitude medical service method according to claim 7, wherein the large model training and service platform development are performed in S2, and specifically comprising the steps of: s21, constructing a one-stop development environment based on Kubeflow, providing a full-flow tool chain from data uploading to model deployment, supporting data set version management, directly configuring a built-in pre-training model, providing real-time dynamic training monitoring, automatically performing model evaluation and optimization after training is finished, and performing model deployment in three modes of cloud-side-end; s22, aiming at a medical unmanned aerial vehicle scene, a special model development tool chain is provided, a multi-task medical unmanned aerial vehicle special model is developed, model training and deployment can be completed only by replacing a data set, and a model development period is greatly shortened; S23, constructing a model warehouse, storing all trained models, supporting model retrieval, model putting-off and model updating, monitoring key indexes of model service in real time, and automatically alarming when the key indexes exceed a threshold value.
10. The ai+ unmanned aerial vehicle-based low-altitude medical service method according to claim 7, wherein the construction of the high-quality medical low-altitude data set in S3 specifically comprises the steps of: S31, setting up an AI low-altitude data labeling platform, completing core labeling work by adopting an AI automatic labeling model, realizing 2D image target frame/category labeling by adopting a YOLOv model, processing point cloud semantics by using a PointNet model, and labeling video frame key frames by using an AI model through time sequence detection; S32, adopting Ceph distributed storage data, adopting 'copy and erasure code' mixed redundancy in a storage system, adopting 3 copies in hot data, adopting 2 copies in warm data, adopting erasure code in cold data, reducing storage cost; S33, a medical simulation scene engine is constructed based on Unreal Engine, building, environment and dynamic element scenes are modeled based on BIM data, digital twin bodies of a plurality of hospital areas are constructed, then three types of data of extreme scenes, abnormal behaviors and multi-mode fusion are generated, the generated data are exported to be in a common format, and labeling information is automatically added and directly used for model training, so that the manual labeling cost is reduced.

Description

Low-altitude medical service system and method based on AI+unmanned aerial vehicle Technical Field The invention belongs to the technical field of intelligent medical services, and particularly relates to a low-altitude medical service system and method based on an AI+unmanned aerial vehicle. Background Medical logistics refers to the process of checking, storing, sorting, distributing and other operation processes in the links of medicine supply, distribution and transportation by effectively integrating the upstream and downstream resources of marketing channels by means of certain logistics equipment, technology and logistics management information systems, improves the order processing capability, reduces goods sorting errors, shortens the stock and distribution time, reduces logistics cost, improves service level and fund use benefits, and achieves automation, informatization and benefit. The low-altitude economy is a novel comprehensive economy form, takes low-altitude flight activities as cores, and takes new mass productivity formed by technologies such as unmanned or unmanned flight, low-altitude intelligent networking and the like to interact with factors such as airspace, market and the like so as to drive the development of low-altitude infrastructure, low-altitude aircraft manufacturing, low-altitude operation service, low-altitude flight guarantee and the like. At present, in the aspect of medical supplies and drug delivery, the traditional conveying mode has high dependence on manpower, and relates to a large number of full-time conveyers and logistics guarantees, so that the labor cost is high and the scheduling coordination is complex. The full-flow informatization level of transportation is not high, digital monitoring and automatic early warning are lacked, the transportation state is difficult to be visible in real time, and the overall logistics management quality improvement and efficiency improvement capability of a hospital is affected. These short boards in traditional mode have been difficult to meet the increasing intelligent management and high aging demands of modern hospitals. Meanwhile, in the aspect of combining low-altitude economic comprehensive application, the problems of difficulty in multi-aircraft coordination, insufficient flight stability of an unmanned aerial vehicle, low-altitude flight airspace conflict, cloud edge end cracking, low resource utilization rate, slow response of emergency tasks and the like in a complex dynamic environment are also faced. In order to solve the problems, the invention provides a low-altitude medical service system and a method based on an AI+unmanned aerial vehicle, which are used for deeply integrating artificial intelligence (large model technology), high-performance computing power and an unmanned aerial vehicle platform, providing an intelligent air service network for medical material distribution, intelligent security inspection, medical emergency response and environment monitoring and killing, comprehensively improving the operation efficiency of hospitals, the security level and the service quality of patients, realizing the crossing from 'perception execution' to 'cognitive decision', breaking through the low-altitude cooperative technical barriers of a cross-park, deeply integrating leading edge technologies such as digital twinning, group intelligence, general sense integration and the like, constructing an intelligent unmanned aerial vehicle service network supporting multi-hospital area linkage, systematically solving three major pain points of medical scene beyond-vision distance scheduling, heterogeneous environment adaptability and cross-domain security management and control, and realizing the omnibearing upgrading of medical resource intelligent allocation. The information disclosed in this background section is only for enhancement of understanding of the general background of the invention and should not be taken as an acknowledgement or any form of suggestion that this information forms the prior art already known to a person of ordinary skill in the art. Disclosure of Invention The invention aims to provide a low-altitude medical service system and method based on an AI+unmanned aerial vehicle, which are used for constructing an industry-leading 'air-ground integrated' medical intelligent logistics network, providing rapid point-to-point distribution of medical drugs, comprehensive dynamic monitoring of a hospital environment and real-time early warning of security risks, and realizing comprehensive upgrading of medical resource intelligent allocation. In order to achieve the above object, the present invention provides the following technical solutions: A low-altitude medical service system based on ai+unmanned aerial vehicle, comprising: the medical transportation unmanned aerial vehicle is used for accurately distributing medical first-aid articles and fire-fighting emergency articles; The intelligent low-alt