Browse Prior Art Database

Power Prediction Method and System

IP.com Disclosure Number: IPCOM000241452D
Publication Date: 2015-Apr-30
Document File: 2 page(s) / 72K

Publishing Venue

The IP.com Prior Art Database

Abstract

Disclosed is a method to predict power demand of buildings without representative statistical method by using large volumes of actual data.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 74% of the total text.

Page 01 of 2

Power Prediction Method and System

Disclosed is a method to predict power demand of buildings without representative statistical method by using large volumes of actual data.

A representative statistical method can predict power demand of buildings, however it will need large volumes of actual data such as past several months or years data to construct a prediction model for future's power demand.

The following method and formula can predict power demand of buildings easily without representative statistical method by using large volumes of actual data.


[1] Categorize a data into a workday and a holiday.

[2] Prepare the latest and past one cycle actual data of workday and holiday.
[3] Predict a power demand by using the formula in Fig.1.

Fig.1

In the formula, "workday_flag" and "holiday_flag" set a value, 0 or 1, based on the category which is predicted day condition (Fig.2). If it is a workday, "workday_flag" sets 1 and "holiday_flag" sets 0. And if it is a holiday, "workday_flag" sets 0 and "holiday_flag" sets 1.

Fig.2

Regarding "workday_past_data", it sets the latest workday's data that is available for use. For example, in the case of "tomorrow's power demand prediction by using up until yesterday's actual data", and if tomorrow is a workday and yesterday was a workday, it sets yesterday's data. (Fig.3)

Fig.3

Regarding "holiday_past_data", it sets the latest holiday's data that is available for use. For example, in the case of "tomorrow's power demand predict...