Analysis of The Development Trend of Domestic Pumped-hydro Storage Based on SVM Model

As the most mature power system regulation device in the current energy storage technology, with the most significant benefit of carbon emission reduction in the whole life cycle, the best economy and large-scale development conditions, pumped-hydro storage plays an important role in ensuring the safety of large power grid, promoting the consumption of new energy and improving the performance of power system. The development of pumped-hydro storage investment is closely related to the geographical conditions, peak shaving demand, renewable energy consumption and so on of the region itself. The non-linear characteristics among the influencing factors are obvious. In this paper, the support vector machine (SVM) is used to establish the prediction analysis model of pumped-hydro storage installed capacity, which is commonly used in machine learning methods. Meanwhile, this model combines the characteristics of different stages of new power system construction, and introduce penalties for external factors such as construction difficulty, investment cost, investment, etc. To adapt to the non-linear characteristics of different influencing factors, the kernel function of this model is a linear combination of polynomial kernel function, RBF kernel function and Sigmoid kernel function. The calculation and analysis results show that the installation rate of pumped storage is expected to peak in 2030, after which some pumped storage devices and micro pumped storage that meet the concept of generalized pumped storage gradually participate in the market competition until 2060, when the era of carbon neutrality combined with electricity and hydrogen comes, pumped storage will not be competitive in the energy storage market.