INTELLIGENT SOLAR-POWERED DIRECT AIR CAPTURE AND ELECTROCHEMICAL CARBON UTILIZATION: A MACHINE LEARNINGENHANCED MULTI-OBJECTIVE OPTIMIZATION AND SLIDING MODE CONTROL FRAMEWORK FOR NET-ZERO INDUSTRIAL DECARBONIZATION
知能型太陽光直接空気回収と電気化学的炭素利用：機械学習強化多目的最適化とスライディングモード制御フレームワークによるネットゼロ産業脱炭素 (AI 翻訳)
Adel Elgammal
🤖 gxceed AI 要約
日本語
本研究は、太陽光発電による直接空気回収（DAC）と電気化学的炭素利用を組み合わせたシステムに対し、機械学習（ニューラルネットワーク、深層強化学習）を活用した多目的最適化とロバストスライディングモード制御を統合した知的制御フレームワークを提案。10kWパイロット試験で従来比23%の効率向上、弱光下でも85%以上の回収率、電圧変動67%低減などを実証。20年間の経済分析では炭素回収コスト31%削減、NPV45%向上、1MW商用システムで年間18万ドルのコスト削減を見込む。
English
This study proposes an intelligent control framework integrating machine learning-based multi-objective optimization and robust sliding mode control for a solar-powered direct air capture (DAC) and electrochemical carbon utilization system. Using a 10kW pilot system, it achieved 23% higher efficiency, >85% capture rate under weak sunlight, 67% lower voltage fluctuations, and <3% conversion deviation versus conventional PI control. Economic analysis over 20 years shows 31% lower levelized cost of carbon capture, 45% higher NPV, and annual savings of ~$180,000 for a 1MW commercial system.
Unofficial AI-generated summary based on the public title and abstract. Not an official translation.
📝 gxceed 編集解説 — Why this matters
日本のGX文脈において
日本は水素・アンモニアやCCUSをエネルギー基本計画に位置づけ、DAC技術の実証も進んでいる。本研究成果は、不安定な太陽光下でも高効率を維持する制御手法を提供する点で、日本の再生可能エネルギー併設型DAC実装に実用的な知見を与える。
In the global GX context
Globally, DAC and electrochemical utilization are key to net-zero industrial emissions but face intermittency and cost challenges. This paper's ML-enhanced hierarchical control framework (optimization + robust control + real-time adaptation) demonstrates significant efficiency and economic improvements, offering a scalable pathway for integrating variable renewables with carbon capture systems.
👥 読者別の含意
🔬研究者:Integrates ML multi-objective optimization with sliding mode control for DAC+electrochemical systems, with experimental validation and economic analysis.
🏢実務担当者:Demonstrates 31% cost reduction and 45% NPV improvement for solar-powered carbon capture, with real-time control adapting to weather and market conditions.
🏛政策担当者:Provides evidence that solar-powered DAC can be economically viable with AI-based control, supporting policy incentives for integrated renewable-CCUS systems.
📄 Abstract（原文）
The urgent need to achieve net-zero decarbonization in the industrial sector is accelerating the research and development of innovative renewable energy-powered carbon capture and utilization technologies. Among these, the pathway that combines direct air capture (DAC) with electrochemical carbon utilization can not only deliver negative emissions, but also produce high-value chemicals simultaneously, making it a highly promising technical direction at present. However, when this energy-intensive process is integrated with intermittent solar energy, it faces three core challenges related to system operation efficiency, working condition stability, and economic feasibility. To address these issues, this study proposes a new intelligent control framework that integrates machine learningaugmented multi-objective optimization and robust sliding mode control. This framework adopts a hierarchical structure: a neural network multi-objective optimization module dynamically balances four core goals, namely maximizing energy efficiency, raising the carbon capture rate, increasing product output, and minimizing economic costs; a robust sliding mode controller ensures stable system operation under variable sunlight and fluctuating loads; and a supporting deep reinforcement learning module adjusts operating parameters in real time based on weather forecasts, energy demand, and market prices. This study completed simulation and experimental verification on a 10kW pilot-scale test system. Compared with the traditional PI control scheme, the overall system efficiency increased by 23%, the carbon capture rate stayed above 85% in weak-sunlight environments, voltage fluctuations dropped by 67%, and the deviation in electrochemical conversion rate was less than 3%. The economic analysis of the full 20-year operation cycle shows that the levelized cost of carbon capture decreased by 31%, the net present value rose by 45%, and deploying a 1MW commercial system can achieve annual cost savings of approximately 180,000 US dollars. This paper sorts out the three core academic contributions and two research limitations to be optimized of this study, fully presenting the core value of this original research and the room for subsequent improvement. This study proposes that economically viable solar-powered carbon capture systems can be widely deployed across industrial scenarios, while also accelerating the implementation of climate targets, creating new economic opportunities in the carbon utilization sector, and advancing sustainable industrial transformation and environmental governance.
🔗 Provenance — このレコードを発見したソース
- openalex https://doi.org/10.33564/ijeast.2026.v11i02.018first seen 2026-06-27 04:55:27
🔔 こうした論文の新着を逃したくない方は キーワードアラート に登録（無料・3キーワードまで）。
gxceed は公開メタデータに基づく研究支援データセットです。要約・翻訳・解説は AI 支援で生成されています。 最終的な解釈・検証は利用者が原典資料に基づいて行うことを前提とします。