At present , the main application scenario of power prediction of new energy is in power planning and dispatching . With the advancement of power reform and the gradual establishment of power trading markets, power forecasting technology will also play an important role in power trading . As the topic of power trading is relatively large, here we mainly stand from the perspective of new energy power stations and analyze the role of power prediction in power spot trading.
What is the spot market?
According to the time scale, the electricity trading market is mainly divided into the medium and long-term market and the spot market . Among them , the spot market refers to the trading market for auxiliary services such as day-ahead, intra-day, and real-time power and reserve, which can be subdivided into day-ahead , intra-day and real-time markets according to the time scale .
In relatively mature electricity markets in Europe and America, the spot market is the core link of the electricity market system. Compared with direct transactions and mid- to long-term transactions for large users with strong planned attributes, the spot market can reflect the supply and demand conditions, power generation costs, etc. in different periods and regions, so as to better discover price information, optimize resource allocation, and improve power system operation effectiveness.
Due to the complexity of my country's electricity market reform, the electricity market that is gradually being established is basically dominated by medium and long-term transactions, supplemented by spot transactions. However, I believe that the importance of spot trading will continue to increase in the future.
The role of power forecasting in the spot market?
For new energy power stations, the current main participation method is "quotation without price guarantee". In the future, it will gradually transition to the method of "quotationquotation". Here we analyze it in two scenarios.
1) " Quotation is not quoted "
If the new energy plant only needs to declare the power generation, this mode is similar to the original power forecast to the dispatcher. The key lies in the accuracy of the forecast. If the accuracy rate is not up to the standard, the power station will be assessed.
The difference lies in the assessment method. For example, in the deviation assessment of the Gansu power market, the actual output will be assessed if the actual output is lower than the planned value by more than -10%, and the actual output will be assessed if the actual output exceeds the planned value by more than 2%. This is different from the previous assessment methods such as the root mean square error or mean absolute error of the same effect of positive and negative deviations.
2) "Quote Quotation"
Compared with "quotation without quotation", "quotation quotation" needs to consider more factors. Here we mainly predict the role of power.
New energy probability prediction will play an important role in power trading. Compared with traditional point prediction, probability prediction provides uncertain information. According to the uncertainty of the forecast, the quotation strategy can be adjusted. If the uncertainty is large, the quotation can be more aggressive; if the uncertainty is small, the quotation should be more conservative.
Through regional power forecasting, you can understand the changes in power generation capacity in the market. Combined with regional power forecasting and load forecasting, you can roughly determine the supply and demand of the entire market, especially in areas with a high proportion of wind power and photovoltaic installations. The role of new energy regional power forecasting is particularly important. Regional power forecasting is an important input for electricity price forecasting . Under the same power demand conditions, the price is low if the amount of electricity is large, and the price is high if the amount of electricity is small.
Conclusion
Today we mainly analyzed the value of power prediction technology for new energy plants in the scenario of power spot trading. Other subjects in the electricity trading market, including traditional power generation companies such as thermal power and hydropower, electricity sales companies, and large power users, also require power forecasting. The so-called "knowing yourself and the enemy will never end in battle."
From a technical point of view, weather forecasting, feature engineering methods, machine learning and deep learning technologies, data transmission schemes, etc. used in power forecasting can be applied to load forecasting and electricity price forecasting in power trading. The so-called "stone from other mountains can be used for jade".
This is the second part of the series of power forecast development trends. I look forward to discussing with you. Finally, I will make a small advertisement. Beijing Xiangxinli Technology focuses on the professional services of electric power weather. Welcome everyone to consult business.
Power forecast development trend series:
"Probability Forecast of Power Forecast Development Trend"
"Electricity Spot Trading on Power Forecast Development Trend"