Research On Human-computer Interaction Technology In The Smart Home

With the advent of the Internet of Things era,smart home has always been a research hotspot at home and abroad.People want to transform their families into a smart,humane personal service place,especially for groups that lack autonomy and independence in life.However,the smart home has not been widely popularized currently because of its replacement and modification of traditional homes which people are already familiar with.The cost of the replacing and the learning are large,so its high intelligence is not enough to overcome its practicality.In order to speed up the arrival of smart home,researchers mostly focus on the design and control of new smart home appliances,and weaken the research on the modification of traditional home appliances or the transition from traditional home to smart home.Therefore,the modification and intelligent control of traditional home appliances are focused on in this paper,using the current new human-computer interaction technology to carry out low-cost modification of home appliances.And the system is compatible with the traditional control methods,so that users can still maintain their original lifestyle.The addition of new control methods makes people gradually realize the convenience that human-computer interaction technology brought to the life,and seek more markets for the smart home.Aiming at the interactive control of traditional household appliances,we investigate and analyze the existing interactive technology schemes and communication implementation methods,and compare their advantages,disadvantages and the scope of application.Finally,we design a set of household appliances control system based on user gesture classification.The system is mainly composed of gesture controller,home appliance receiver and central processing unit,using ZigBee communication for system data transmission.The gesture controller is responsible for the positioning of the home appliance,and connects with the target through infrared communication.Its sensor collects the user's arm acceleration and uploads it to the central processor for gesture classification and discrimination.The processor transmits the classification result to the receiver and makes it control the appliance specifically.In view of the current mainstream control mode and wiring characteristics of traditional home appliances,we design three receiver schemes suitable for different types of home appliances,including the type of socket,the type of infrared remote control and the type of single-wire power switch,and add gesture control function on the basis of compatibility with traditional control mode.We also design a more general infrared learning decoding and coding scheme,because the protocol of infrared remote control is not consistent.We choose four simple dynamic gestures for the control mode of household appliance system,which are suitable for the functional characteristics of home appliances and human body control habits.The common methods of data preprocessing and classification of gestures are analyzed in this paper,and we choose the method of smooth filtering and energy extremum for data smoothing and the interception of effective data.Principal component analysis is used to analyze the dimensionality reduction of the four gesture data.We use the linear classification method to classify the four gesture data with two-dimensional features,and the classification accuracy is up to 98%.