提出一种与质量相关的核潜结构投影方法,并将其应用于非线性过程的质量监测中.结果表明,与KPLS方法在过程潜变量空间监测质量相关的变化不同,新方法通过将质量变量空间进一步分为3个子空间,实现了对包含能被过程解释的质量部分以及不能被过程解释的质量部分的完整质量监测.同时,通过TE仿真案例验证了新方法的有效性.
A novel quality-related kernel latent structure projection method was proposed and quality-relat- ed monitoring was applied in the nonlinear process. Different to the KPLS, the proposed method further decomposes the quality variable space into three subspaces, which can achieve a complete monitoring of the quality which can be and cannot be interpreted by process quality. Meanwhile, the validity of the proposed method was verified by TE simulation.