Study on purification system of air separation equipment based on partial least square method

The predictive control of purification system of air separation plant was researched based on the modified orthogonal signal of partial least squares method. It will analyze several defects of air separation plant and purification systems, the short switching time can reduce the useful life, the long time of the regeneration can lead large consumption of the impure nitrogen and so forth. A least squares method regression model of controlling the switching time of molecular sieve was established by using the experimental data of the purification system. And the model improved the system performance and reduced the computation using the modified orthogonal signal method. The validity of the model excellent agreed with the experimental result.