Penelitian ini bertujuan untuk memodelkan dan mengoptimasi sistem terintegrasi berbasis energi terbarukan untuk produksi hidrogen hijau dan penangkapan karbon dari udara menggunakan pendekatan Reinforcement Learning. Sistem terdiri dari empat komponen utama: Parabolic Trough Solar Collector (PTSC) sebagai sumber panas, Rankine Cycle sebagai pembangkit listrik, Proton Exchange Membrane Electrolyzer (PEME) untuk elektrolisis air menjadi hidrogen, dan Direct Air Capture (DAC) untuk menyerap CO₂ dari atmosfer. Pemodelan dan simulasi dilakukan menggunakan Engineering Equation Solver (EES), sedangkan proses optimasi dijalankan dengan Python melalui kombinasi Artificial Neural Network (ANN), Multi-Objective Genetic Algorithm (MOGA), dan Multi-Objective Reinforcement Learning (MORL). Validasi model menunjukkan tingkat kesalahan rata-rata rendah, yaitu 4%. Berdasarkan simulasi, sistem menghasilkan daya bersih 262,4 kW, laju produksi hidrogen 0,0001605 kg/s, dan penangkapan CO₂ sebesar 0,1075 kg/s, dengan efisiensi eksergi total 7,96% serta biaya produksi hidrogen 55,42 $/GJ. Analisis sensitivitas menunjukkan bahwa peningkatan radiasi matahari, tekanan masuk turbin, dan laju aliran massa PTC secara signifikan meningkatkan efisiensi eksergi dan laju produksi hidrogen, serta menurunkan biaya sistem. Hasil optimasi memetakan Pareto Front yang menunjukkan hubungan sinergis antara efisiensi dan produksi hidrogen, serta trade-off terhadap biaya produksi. Penelitian ini diharapkan menjadi referensi dalam pengembangan sistem energi bersih yang efisien dan ekonomis.

---

This research aims to model and optimize an integrated renewable energy-based system for green hydrogen production and atmospheric carbon capture using a Reinforcement Learning approach. The system consists of four main components: a Parabolic Trough Solar Collector (PTSC) as the heat source, a Rankine Cycle for electricity generation, a Proton Exchange Membrane Electrolyzer (PEME) for water electrolysis to produce hydrogen, and a Direct Air Capture (DAC) unit for CO₂ removal from the atmosphere. Modeling and simulation were performed using Engineering Equation Solver (EES), while the optimization process was conducted in Python through a combination of Artificial Neural Network (ANN), Multi-Objective Genetic Algorithm (MOGA), and Multi-Objective Reinforcement Learning (MORL). Model validation showed low average errors, namely 4%. Under base case conditions, the system produced a net power output of 262.4 kW, a hydrogen production rate of 0.0001605 kg/s, and a CO₂ capture rate of 0.1075 kg/s, with an overall exergy efficiency of 7.96% and a hydrogen production cost of 55.42 $/GJ. Sensitivity analysis indicated that increasing solar radiation, turbine inlet pressure, and PTSC mass flow rate significantly improved exergy efficiency and hydrogen production while reducing system cost. Optimization results mapped a Pareto Front highlighting the synergistic relationship between efficiency and hydrogen output, as well as trade-offs with production cost. This study is expected to serve as a reference for developing efficient and economical clean energy systems.