Policy scenario analysis for solar-wind decarbonized generation expansion planning using AI-based load forecasting and adaptive hybrid Aquila -Harris Hawk Optimization

• Hybrid LSTM–XGBoost model for accurate short- and long-term load forecasting. • Adaptive Hybrid Aquila Optimizer–Harris Hawk Optimization for GEP decision-making. • Policy-oriented generation expansion planning with solar, wind and energy storage. • Multi-objective optimization of system cost, emissions, reliability and uncertainty. • Cost and reliability assessment under varying forced outage rates of renewables. Power system generation expansion planning plays a vital role in achieving net-zero carbon emissions and ensuring a sustainable energy future, also providing a reliable power supply to end-users. The study presents a hybrid artificial intelligence (AI) − based load forecasting model and constructs a generation expansion planning (GEP) model using a hybrid metaheuristic optimization algorithm. In this study, a hybrid Long Short-Term Memory (LSTM)- Extreme Gradient Boosting (XGBoost) algorithm is proposed for the load forecasting analysis, which shows better predictive results based on performance evaluation metrics. The model achieves low error metric values, such as mean absolute error (MAE) of 0.0340 and 0.2804, mean squared error (MSE) of 0.0022 and 0.1491, and root mean squared error (RMSE) of 0.0472 and 0.3850 for both training and testing, along with high prediction accuracies of 99.88% (training) and 98.96% (testing). A novel Adaptive Hybrid Aquila Optimizer-Harris Hawk Optimization algorithm (AHAOHHO) is introduced to support capacity planning over 6-year and 14-year transition horizons. This optimization algorithm aims to reduce total system costs, uncertainty, and environmental effects, while enabling policymakers to assess alternative policy strategies for effective integration of renewable energy resources. The outcome of this study presents a cost comparison analysis based on different forced outage rates (FORs) of solar and wind capacity, revealing a clear economic trend in all policy scenarios.