A risk-based optimization framework for integrated supply chains using genetic algorithm and artificial neural networks

Modern supply chains are complicated networks that stretch over different geographical locations, which make them vulnerable to a variety of risks. An effective supply chain management provides a competitive advantage by reducing overhead costs and delays in product deliveries. To cope with internal and external risks, an integrated planning scheme can effectively adjust supply chain network operations and minimize negative effects of unexpected changes and failures. This paper proposes a risk-based optimization framework that handles supply chain's strategic, tactical, and operational decisions. A supply chain is considered as a network of suppliers, manufacturing plants, distribution centers, and markets. The proposed model deals with uncertainties associated with demands, facility interruptions, lead times, and failures in supply, production, and distribution channels. To study more realistic supply chain operations, delays uncertainties are also included in the model. The structural design, communication between different centers, and inventory decisions are determined based on the risk perspective of the decision maker. To solve the proposed model, a new genetic algorithm is designed which is integrated with artificial neural network that learns from previous plans and search for better ones by minimizing any mismatch between supply and demand. The effectiveness of the proposed framework is investigated by comparing its results with those obtained from traditional techniques and regular GA. The results show that including adjustable tactical plans and incorporating learning mechanism considerably reduces inventory level and increases the profit level in supply chain systems.