Research On Medical Waste Tracing System Based On Colored Petri Net And Deep Learning

The outbreak of covid-19 has led to a surge in confirmed cases of new crown pneumonia,while the number of medical waste has also been increasing in a spurt of growth.The traditional written record method of medical waste management has some problems,such as low efficiency,difficult tracing of medical waste,easy contact and infection of staff,which has been difficult to adapt to the current management of medical waste.In addition,medical waste is highly polluting and hazardous.If medical waste is not strictly supervised and the trend of medical waste management is traced in time,once medical waste is lost,leaked or improperly disposed,It will seriously affect people’s life and health,and then lead to a series of economic and social problems.Therefore,based on colored Petri net and deep learning theory,this paper constructs a set of medical waste tracking system to solve the above problems.The main research contents of this paper are as follows:(1)The workflow of medical waste management inside and outside medical institutions is analyzed,and the medical waste traceability and tracking system model is constructed based on colored Petri net.The model can realize the traceability and tracking management of medical waste from generation to destruction.In this paper,the model is simulated and verified based on CPN tools modeling tool,and the dynamic properties of the system model are analyzed by using incidence matrix.The results show that the established colored Petri net medical waste traceability and tracking system model has accessibility,boundedness and activity,and can realize the online traceability and tracking management of medical waste.(2)The established colored Petri net medical waste tracking system model realizes the information tracking of medical waste in bag or box,but it can’t avoid the potential safety hazard of medical waste leakage caused by the damage of medical waste packaging bag.Therefore,this paper proposes a medical waste packaging bag damage detection algorithm based on improved Yolov5.The algorithm combines the shallow features with the deep features,and establishes the fourth target detection layer to improve the detection accuracy of small damaged targets of medical waste packaging bags;For the medical waste packaging bag data set,the K-means clustering algorithm is used to obtain a new a priori frame,and the convolutional block attention module integrating space and channel is introduced,so that the network can better focus on the useful features of the target and ignore the useless features.The experimental results show that compared with the original network,the average detection accuracy map of the improved Yolov5 network is improved from 88.6% to 94.1%.Compared with SSD,Yolov3 and fast R-CNN,the improved network has higher detection accuracy and faster reasoning speed in the task of medical waste packaging bag damage detection.(3)Based on the constructed colored Petri net medical waste tracking system model and the improved medical waste packaging bag damage detection model of Yolov5,the medical waste tracking system is designed and implemented.It realizes the mobile APP and web system based on android Studio,Intelli J IDEA and Pytorch development platform.App can submit and store the data of all links of medical waste management to the cloud database,and the web system can integrate the data submitted by APP,so that relevant supervisors can control the medical waste information from multiple dimensions,realize the information traceability management of the whole life cycle of medical waste,automatically detect the damaged medical waste packaging bags,and further eliminate the potential safety hazards of medical waste leakage caused by the damage of medical waste packaging bags.There are 116 figures,33 tables and 53 references.