Project finance is used in capital intensive renewable energy projects worldwide. Financial entities such as large banks and institutional investors are active in providing syndicated loans for infrastructure projects and compete to offer better terms to the sponsors of these projects. The literature is full of research on capital structure optimization. We propose a novel stochastic framework for optimizing the debt service schedule with due regard to the probability of default of the project company. The applicability of the proposed method is demonstrated for a 10 MW solar photovoltaic project employing a genetic algorithm (GA) as the optimization tool. The NASA SSE dataset is used to collect irradiation data, and PVsyst software is employed to simulate the performance of the project. The uncertainties are accounted for using Monte Carlo simulation, and the revenue generated, its corresponding free cash flow and the debt service coverage ratio are simulated as random variables. The proposed optimization framework enables lenders to offer an optimized debt service that maximizes the shareholder's profitability index. A particle swarm optimization is also employed to validate the stability and usefulness of GA optimization.

项目融资被用于全球资本密集型可再生能源项目。大型银行和机构投资者等金融实体积极为基础设施项目提供银团贷款，并竞争为这些项目的发起人提供更好的条款。文献中充斥着关于资本结构优化的研究。我们提出了一种新颖的随机框架，用于在适当考虑项目公司违约概率的情况下优化债务偿还时间表。利用遗传算法（GA）作为优化工具，证明了该方法在10 MW太阳能光伏项目中的适用性。NASA SSE数据集用于收集辐照数据，PVsyst软件用于模拟项目的绩效。不确定性使用蒙特卡洛模拟法解决，并将产生的收入，其相应的自由现金流量和偿债率作为随机变量进行模拟。拟议的优化框架使贷方可以提供优化的债务服务，以最大化股东的盈利能力指数。粒子群优化算法也被用来验证遗传算法优化算法的稳定性和实用性。