Scholarlink

This study proposes an adaptive weight multi-objective model predictive control (AW-MOMPC) framework for batch cooling crystallization process, dynamically adjusting objective weights to balance product quality and energy efficiency. By correlating the weight with real-time supersaturation levels, the algorithm prioritizes crystal growth suppression during early-stage nucleation while minimizing energy consumption in later phases. Compared to linear cooling, the multi-objective model predictive control (MOMPC) strategy reduces fine crystal volume by 24% and energy consumption by 25.6% without compromising product quality. However, fixed-weight MOMPC faces challenges in multi-objective equilibrium, leading to suboptimal performance. In contrast, AW-MOMPC achieves near-identical fine crystal suppression as single-objective MPC but reduces energy consumption by 14.2%. The adaptive mechanism effectively resolves the trade-off between crystal size distribution and operational costs, validated by third-order moment analysis and final crystal density profiles. This work demonstrates the viability of weight-adaptive MPC in enhancing both economic and quality outcomes for complex crystallization systems.

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

Multi-objective Model Predictive Control with Adaptive Weighting for Batch Crystallization Process

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Zimu Diao,
  • Fangkun Zhang,
  • Qilei Xu,
  • Baoming Shan

摘要

This study proposes an adaptive weight multi-objective model predictive control (AW-MOMPC) framework for batch cooling crystallization process, dynamically adjusting objective weights to balance product quality and energy efficiency. By correlating the weight with real-time supersaturation levels, the algorithm prioritizes crystal growth suppression during early-stage nucleation while minimizing energy consumption in later phases. Compared to linear cooling, the multi-objective model predictive control (MOMPC) strategy reduces fine crystal volume by 24% and energy consumption by 25.6% without compromising product quality. However, fixed-weight MOMPC faces challenges in multi-objective equilibrium, leading to suboptimal performance. In contrast, AW-MOMPC achieves near-identical fine crystal suppression as single-objective MPC but reduces energy consumption by 14.2%. The adaptive mechanism effectively resolves the trade-off between crystal size distribution and operational costs, validated by third-order moment analysis and final crystal density profiles. This work demonstrates the viability of weight-adaptive MPC in enhancing both economic and quality outcomes for complex crystallization systems.