Scholarlink

<p>Electric vehicles (EVs) are pivotal to achieving planetary sustainability, with nickel playing a critical role in high-performance battery technologies such as Nickel-Manganese-Cobalt (NMC). As global nickel demand surges, laterite ores have become the primary source, necessitating efficient extraction methods like high-pressure acid leaching (HPAL). However, optimizing high-pressure acid leaching for maximum nickel recovery and minimal environmental impact remains challenging due to the complex interplay of process parameters. This study uniquely combines a systematic comparison of eight machine learning (ML) models—Linear Regression, Ridge Regression, Lasso Regression, Random Forest, Extreme Gradient Boosting, Neural Networks, Support Vector Machine, and Generalized Additive Models—with response surface methodology (RSM) to predict and optimize nickel extraction efficiency from laterite ores using high-pressure acid leaching. Response surface methodology was employed to design experiments and analyze interactions between key variables, such as temperature, acid concentration, and leaching time. Model performance was assessed using statistical metrics, and factor interactions were visualized through 3D response surface plots. Shapley additive explanations (SHAP) were applied to enhance model interpretability, revealing the relative importance of each parameter. The results demonstrate that machine learning, specifically Neural Networks model combined with response surface methodology can effectively optimize high-pressure acid leaching processes, facilitating accurate representation of the complex interactions between parameters, and reducing reliance on extensive experimental trials. This approach not only identifies optimal parameter combinations for maximizing nickel recovery but also minimizes environmental risks. By advancing sustainable nickel extraction, this study supports the growing demand for nickel in clean energy technologies and contributes to the global transition toward electric mobility.</p> Graphical Abstract <p></p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Enhancing Nickel Recovery from Laterite Ores via High-Pressure Acid Leaching: Machine Learning and Design of Experiments for Process Optimization and Environmental Sustainability

  • 期刊论文

  • 发表日期：

    2025年10月15日

查看全文
  • Binathara Priagung Surya,
  • Wei Liu,
  • Xu Yan,
  • Kaihua Xu,
  • Qingwei Wang,
  • Feiping Zhao,
  • Liyuan Chai

摘要

Electric vehicles (EVs) are pivotal to achieving planetary sustainability, with nickel playing a critical role in high-performance battery technologies such as Nickel-Manganese-Cobalt (NMC). As global nickel demand surges, laterite ores have become the primary source, necessitating efficient extraction methods like high-pressure acid leaching (HPAL). However, optimizing high-pressure acid leaching for maximum nickel recovery and minimal environmental impact remains challenging due to the complex interplay of process parameters. This study uniquely combines a systematic comparison of eight machine learning (ML) models—Linear Regression, Ridge Regression, Lasso Regression, Random Forest, Extreme Gradient Boosting, Neural Networks, Support Vector Machine, and Generalized Additive Models—with response surface methodology (RSM) to predict and optimize nickel extraction efficiency from laterite ores using high-pressure acid leaching. Response surface methodology was employed to design experiments and analyze interactions between key variables, such as temperature, acid concentration, and leaching time. Model performance was assessed using statistical metrics, and factor interactions were visualized through 3D response surface plots. Shapley additive explanations (SHAP) were applied to enhance model interpretability, revealing the relative importance of each parameter. The results demonstrate that machine learning, specifically Neural Networks model combined with response surface methodology can effectively optimize high-pressure acid leaching processes, facilitating accurate representation of the complex interactions between parameters, and reducing reliance on extensive experimental trials. This approach not only identifies optimal parameter combinations for maximizing nickel recovery but also minimizes environmental risks. By advancing sustainable nickel extraction, this study supports the growing demand for nickel in clean energy technologies and contributes to the global transition toward electric mobility.

Graphical Abstract