This paper proposes a stochastic optimal technique based on Genetic Algorithm (GA) for the calculation of multiphase and multicomponent chemical equilibrium by minimization of Gibbs free energy. Three aspects are improved based on the drawbacks of the conventional GA.An alternative decimal encoding strategy is adopted to enhance the precision of calculation.A dynamic encoding method that can limit the bounds of optimized variables within their feasible regions is developed to cope with the complex constraints of the problem.Finally

sequential search technique is applied to improve GA to approach global optima.It is shown through the calculation of complex chemical systems

in which non-ideal

multireaction and multiphase coexistence are simultaneously involved

that the presented GA is general and efficient for the addressed problem.