Research On Assistant Decision System Of Smart Home Based On SVM

With the improvement of living standards,Internet of Things,development of machine learning and other technology,how to make smart home with independent "learning" ability,self-control of lighting,curtains and other home opening and closing according to users' own habits,is becoming one of hot spots in the current smart home research,which has great significance of the smart home's development and promotion.In this paper,we study deep processing of the abnormal data,set of the user decision rule,decision of the user classification and the decision's optimization in the assistant decision-making system of intelligent home.The main work of this paper is as follows:1)Analyze and study the characteristics of environment information,decision-making method and evaluation standard in assistant decision-making system,and give a decision model based on support vector machine and K-nearest neighbor algorithm.This model mainly includes personal environment information modeling,exception data processing,similar user search and user decision rule set acquisition based on SVM.Compared with general decision system and the decision system based on BP neural network in the experiment,the result shows that this method has the characteristics of less time consuming and high precision.2)An improved firefly algorithm based on adaptive step size and enhanced stochastic perturbation is proposed to optimize the decision system.By introducing the luminance value and the Euclidean distance formula,the position updating formula in the original algorithm is improved.The test function in matlab has tested this algorithm,which proves that the algorithm has some advantages in avoiding the problem of premature of convergence and unable to convergence.3)Design and implement of the assistant decision-making prototype system of intelligent home.The research contents include the overall architecture design of smart home system,the collection of personal environmental information and the design and development of smart home APP for viewing and controlling the home.