Design Of Medical Waste Identification And Sorting Bin

Medical institutions generally have problems such as untimely and inaccurate classification of medical waste,outflow of medical waste,and mixing of garbage.Existing medical trash cans have shortcomings such as strong circulation,low intelligence,and professional functionality.The timely and accurate classification of medical waste is an important part of improving hospital feelings.With the intelligent classification of medical waste bins as the carrier,we will promote the standardized treatment of medical waste from the source,control the outflow of medical waste,and make up for the shortcomings of medical waste disposal.There is no room for delay.The research and development of existing smart sorting trash bins at home and abroad is still at the stage of identifying and sorting individual waste.Most of the products that classify individual garbage are still in the realization of the recognition function,requiring users to wait for the results and deliver them independently,and the degree of automation is low.Smart trash cans with automatic sorting function have problems such as long sorting time,high energy consumption,and low space utilization.The main research problem of this paper is the multi-target recognition of smart waste sorting products,with the purpose of developing flexible,efficient,low-energy-consumption medical waste sorting products.The research content of this paper is as follows:(1)The overall architecture of the product was designed,and the size and main functional areas were determined based on the analysis of user needs.It was designed according to the using environment,functional characteristics,ergonomics,mechanical knowledge of industrial design.(2)The Delta sorting mechanism analyzed and reversed its working space.The inverse kinematics model of Delta mechanism was established.The selection of motor and the modeling of sorting mechanism were carried out.The parts of sorting mechanism were designed by Solid Works software.The assembly of hardware and mechanism motion were simulated.Finally,some parts were obtained by 3D printing for hardware assembly.(3)To develop the control program of sorting mechanism,the calculation of inverse kinematics solution and the execution of end positions were converted into motor rotation steps.The method was written as the interface to receive the target category and position information output by the identification system.The sorting path was planned,and then the target was put into the corresponding dustbin after identification and grabbing.(4)The sample collection environment was set with collection equipment.The medical waste sample images were collected,filtered and labeled.The original images were mapped to establish the training sample set.Then the target detection and recognition model was trained by YOLOv4,and was connected with the motor control system.