In supply chains, the phenomenon of bullwhip effect (the variance amplification of order quantities observed in supply chain) has received a considerable attention by companies as it leads to tremendous losses and poor customer services. This variance amplification occurs according to the necessity of using forecasting methods by companies to predict the demand. To overcome this problem the business society resorts to develop what is called decision rules which show that the bullwhip effect is not avoidable. This paper introduces a control engineering approach called Derivative Derived Generalized Predictive Control (DDGPC) together with Genetic Algorithm GA to reduce this effect. The proposed method can reduce bullwhip effect. Moreover, stability and robustness analyses of the proposed technique are investigated. This would help decision-makers in supply chains management to reduce the negative consequences of the bullwhip effect.