Novel Feasible Path Optimisation Algorithms using Steady-State and/or Pseudo-Transient Simulations

The feasible path optimisation algorithm is widely used for chemical process optimisation due to its effectiveness for large-sale highly nonlinear optimisation problems. The effectiveness strongly relies on the convergence of process simulation as the simulation is performed in each optimisation step. Existing feasible path optimisation algorithms often fail to converge or require expensive computational effort in process simulation in the equation-oriented environment. In this work, we propose three novel feasible path optimisation algorithms to improve both convergence and computational efficiency. The first algorithm improves the original steady-state feasible path algorithm through resetting the initial point used for process simulation when its direct precedence fails during the line search. The second algorithm is an enhancement to the time-relaxation-based optimisation algorithm through the use of the tolerances-relaxation integration method for the pseudo-transient continuation (PTC) simulation. The last algorithm is a hybrid algorithm through the effective combination of steady-state simulation and PTC simulation in the feasible path optimisation framework. The computational results demonstrate that the proposed three new variants do improve the convergence and are more robust than the existing feasible path optimisation algorithms. The first variant is more efficient than the other two and the existing PTC model-based optimisation algorithms with the same computer hardware and software.