AI Time Series Control System Modelling by Chuzo Ninagawa

Author:Chuzo Ninagawa
Language: eng
Format: epub
ISBN: 9789811945946
Publisher: Springer Nature Singapore

4.3 Practical Example 5: Electricity Wholesale Market LSTM Model

4.3.1 Prediction of Wholesale Electricity Prices

In this section, we describe a practical example of trend forecasting that depends on time series history using LSTM neural networks. As an application field of the example, market price trend prediction is considered, which is one of the standard time series trend prediction. In this section, we try to forecast the trend of the hourly price of a newly established wholesale electricity market.

For many years, as an example, Japan’s commercial electric power system has been operated by a single company, known as vertically integrated, from generation to transmission and then to retail. In recent years, policies promoting the separation of power generation and transmission have gradually led to the separation of power generation, transmission, and retail into separate entities.

As an example, the Japan Electric Power Exchange (JEPX), an electricity wholesale market, was established in Japan to provide a mechanism for trading electricity between power generation and retail companies. As of 2020, about 800 million kWh, or about a quarter of Japan’s total daily power generation of just under 3 billion kWh, has been procured from the spot market.

In 2016, a one-hour-ahead trading system was established. In this one-hour-ahead trading, sell and buy orders are placed up to one hour in advance using the Zaraba method (price priority principle and time priority principle) every 30 min throughout the day. Each transaction is executed one after the other in a matching manner [2]. Figure 4.8 shows a time series example of actual transactions in the JEPX 1-h-ahead market on August 6, 2018.

Fig. 4.8Actual trading in the JEPX 1-h premarket (August 6, 2018) (Compiled from the Japan wholesale power exchange website: pre-hourly market trading results) [1]

Download

Copyright Disclaimer:
This site does not store any files on its server. We only index and link to content provided by other sites. Please contact the content providers to delete copyright contents if any and email us, we'll remove relevant links or contents immediately.