Abstract:

In this paper, an adaptive batch optimization method based on T-PLS is proposed to apply to the quality control and optimization of batch process. Due to the complexity of batch process, the model can’t perfectly match with the plant and result to NCO mismatch, in which the optimization result is just suboptimal. The gradient correction method is used to compensate the differences between the model and the plant. At the

same time, in order to expand the effective range of control variables, T-PLS statistics (,,) are

introduced as a hard constraint. Finally, the simulation is carried out to verify the effectiveness of the proposed method.

Key words: Batch process, adaptive optimization, T-PLS, hard constraints

摘要： 提出一种基于(Total Projection to Latent Structure, T-PLS)的批次间自适应优化策略用于解决批次过程质量优化问题。由于批次过程的复杂性，模型和对象之间经常出现不匹配，进而导致了最优解必要条件（NCO）不匹配问题，最终致使优化结果只是次优。使用梯度修正方法弥补模型和对象之间的差异，同时为了保证操控变量在有效区间之内，引入了T-PLS的统计量、、作为硬约束条件。最后通过仿真实验，验证了所提

方法的有效性

关键词: 批次过程, 自适应优化, T-PLS, 硬约束条件