Distribution Network Digitalauxiliary Investment Decision-Making Method Based on Highly Relevant Features

With the rapid development of renewable energy, the distribution network is facing increasing challenges, but the access of renewable energy to the distribution network has also brought new development to the distribution network. Digital investment in distribution networks has certain potential benefits, but traditional investments in power grids are difficult to cope with situations that require rapid decision-making. This paper aims to identify key influencing factors in distribution network digital investment through feature selection methods to improve the efficiency and accuracy of decision-making. This paper first uses the Latin hypercube sampling method to obtain the input sample data set, and calculates the performance target through the traditional distribution network digital investment performance calculation model to form an investment-related data set. Subsequently, the data is z-score standardized and input into the feature selection model to perform feature selection for multiple performance targets and extract high-relevant feature sets. Finally, through the high relevant feature set, direct and targeted guidance is provided to decision makers to assist them in making efficient investment decisions. This paper provides a method for power grid enterprises to respond quickly and make efficient decisions in emergencies.