A sampling-based stochastic optimization for a boiler process in a pulp industry

Our Aim was to find the stochastic optimal solution for a boiler process in order to maximize steam generation while complying with pollutant emissions regulations. For a simulation base-model, support vector regression (SVR) is employed to present a general discrete-time dynamical system of a boiler, and a stochastic optimization was performed to take inherent process uncertainties into account. To generate a stochastic optimization problem, sample average approximation (SAA) based on Monte-Carlo sampling was introduced due to the properties of a SVR. Moreover, a gradient free-based particle swarm optimization (PSO) technique was applied to find the optimal parameters of SVR and investigate the stochastic optimal solution for the boiler process. The results show that the stochastic optimal solution provides almost 2% more steam generation under uncertainties: this indicates that the stochastic optimal solution provides more realistic results compared to the deterministic approach. The proposed methodology can be applied straightforwardly to black box models, or when the use of gradient-based optimization solvers is restricted.