Technology For Forecasting Remaining Bath Time Of Water Heater Based On The Data Driven Method

Rise of the concept of intelligent home appliance reflects the growing psychological needs of users for the traditional electrical appliances intelligence.This concept put forward the new technical requirements for the traditional home appliance industry,at the same time,it pointed out the new innovation space and innovation direction for them.As a part of intelligent home appliance,Bath visualization can show accurate prediction time fleetly,then prevent phenomenon of hot water shortages or surpluses and enhance user bathing experience,it,at the same time,can play heuristic and fundamental role to other research such as optimal installation location of heating pipe.This paper studies the prediction method based on data-driven method applied to problem of water heater residual bath time prediction.The main research contents are summarized as follows:Firstly,heat transfer process of the water heater tank was simulated based on Fluent and selected four position for data acquisition that can represent overall changes of tank temperature.Secondly,pre-processing algorithm was designed according to the structure and characteristics of collected data to clean error data and eliminate data noise.Thirdly,all region data were generated based on characteristics of normal bath data,which is the basis of data fitting and data forecasting.Fourthly,algorithm of data partition was designed based on the accuracy requirement and diversity factor of deadline,so as to insure that best ratio can be found by GBL algorithm.Fifthly,GBL algorithm(GA+BP+LUBE)was designed in view of the forecast requirement that "fixed percentage of the prediction error of predicted data should be in specified error range" to train and forecast data,and its forecasting result and ability to get best value was verified.Sixthly,water heater data forecasting system was designed and realized based on WINFORM.The system includes modules of data management,data forecasting,data analysis,image display and system settings.It can realize technical route of remaining bath time forecasting.