Research on Artificial Intelligence - Assisted Financial Analysis Methods in Microgrid Project Investment Decisions

Facing the globalization energy revolution, this paper proposes an intelligence analyzing system based on artificial neural network(ANN), genetic algorithm(GA) and fuzzy logic to solve the dynamic adapatability lack of the traditional financial analysis for microgrid investing, by setting up a database of more than 500 microgrid historical investing projects, and its financial performance forecasting neural network has achieved 0.0243 mean squared error(MSE).As can be seen, when a 23% improvement has been obtained in the accuracy of prediction between traditional linear regression and our hybrid model at 0243. Experimental results show that GA optimization module can improve the project net present value (NPV) by 6.7%(from ¥1.5 million to ¥1.6 million) within the budget constraint and increase renewable energy penetration to 75%. The fuzzy logic risk assessment model which is based on the 27 expert-defined rules, continues to maintain positive risk-adjusted NPV at ¥1.45 robustness gains can be achieved, with robust decisions found with greater than at 18% higher chance with our framework . The framework captures implicit factors such as equipment efficiency decay and energy price volatility, reducing the payback period from 5 years to 4.8 years and significantly improving portfolio risk resilience. .In this paper, a reasoning decision framework of investment recommendation in microgrid is conducted to realize comprehensible microgrid investment planning and decision-making system based on modular structure can be quickly and conveniently updated for multiple application regions and provide an important guidance for sustainable planning and investment of energy.