IEEE Access (Jan 2022)
Internet of Things Intelligent Interaction Technology Using Deep Learning in Public Interaction Design
Abstract
In order to improve the intelligence and humanization of the home environment, smart air conditioners are used as the research objects. User needs using the theory of Human–Computer Interaction (HCI) are investigated and researched, and smart air conditioners are researched and practiced on human interaction design. Firstly, in order to establish a user model, the questions and needs of the target users are explored through the questionnaire survey method. Secondly, from the perspective of ergonomics, the hardware and software interaction interface of the smart air conditioner is analyzed. Finally, the intelligent air conditioner is designed for HCI according to user needs, and the unsupervised feature value extraction of the vibration measurement signal of the outdoor unit of the household air conditioner is carried out by using the stacked autoencoder neural network. Three preliminary options for the design of remote controllers for air conditioners are proposed, and the fuzzy evaluation method is utilized to analyze and evaluate the three options. The research results show that the comprehensive evaluation results of the three preliminary options of the remote-control design are 0.78, 0.77, and 0.8 respectively. Compared with Option 1 and Option 2, Option 3 has obvious advantages. The design of Option 3 is more prominent in terms of comfort, aesthetics, and rationality. Therefore, Option 3 is selected as the final design solution. Under different hidden layer numbers and node numbers, the classification accuracy rate changes in a convex function. When the number of layers is 3 and the number of nodes is 100, the classification accuracy rate is the highest. According to the needs of users, a specific interactive design analysis is carried out on the hardware design of home appliances from the perspective of ergonomics.
Keywords
