My Interest Research Area
- Big Data Analysis
- Forecasting of Commercial Building Power Consumption
- Sensor Network
- Network Protocols
- Network Processor Architecture
Research Area
After huge earth quake, power becomes very serious matter. Power supply and demand planning are getting tougher and tougher since Fukushima nuclear energy plant has to be shutdown due to enourmous damage. Power consumption forecasting becomes very important for power generation and supply companies to secure power supply to industrial facilities such as manufacturing factories, commercial buildings, and residential homes. My research goal is to build up mathematical model from the University of Tokyo sensor data and implement prediction program which predicts future power consumption of commercial building. To achieve it, we will use general time-series data and its analysis. First of all, we will look the University of Tokyo sensor data. In Green University of Tokyo Project (GUTP), over 2000 sensors are deployed in the university campuses and sensor data is recorded in every minute. These data is stored into database and 1 year data is grown hundreds of giga bytes. By analysing this data, we will get characteristics and make formula to predict it. In the future, we will apply prediction model generation by using machine learning. Our prediction will contribute commercial building power consumption so that building operator can make decision based on the prediction data. Prediction accuracy target is about less than 2% as my research. For any helps or opinions, please send me an e-mail. E-mail address is at contact page.