An integrated solution approach based on interdiction problems forurban water networks

Infrastructures play a critical role to provide some resources such as electricity, water, natural gas etc. to people and ensuring social order and quality of life in societies. However, occasionally, critical infrastructures may face threats such as interruptions due to system failure, cyber-attacks, natural disasters, and terrorist attacks. These threats affect the security and functioning of infrastructures, the economy of countries, and the welfare of societies. Therefore, infrastructures must be flawlessly established, sustainable, resilient, and protected. In this study, urban water networks, which are one of the critical infrastrucures are discussed. Water networks protect public health by providing clean and reliable drinking water to cities and in this study they are analyzed within the scope of an interdiction problem..The three-level (defender, attacker, and operator) mathematical model developed by Wu et al. (2021) was analyzed. Initially, the two-level version, consisting of the defender and attacker, was solved using both a mathematical model and a genetic algorithm and subsequently, the three-level version was solved using a greedy algorithm. Results were examined by generating problems of different scales. According to the solution methodology , first, all protection plans for an urban water network are determined. Then, alternative attack plans are generated corresponding to each protection plan. The attacker tries to maximize the damage as response for the defenders plans. The problem is solved by selecting, for each protection plan, the best attack plan for the attacker (the plan that corresponds to the maximum damage that the attacker can inflict against that protection plan) and the best defense plan for the defender (the least-damaging of the plans, each of which represents the maximum damage that the attacker can inflict against a given defense plan). This approach is a defender-attacker-oriented approach. In this context, budgetary constraints are also taken into account. This study is coded in Python 3.11.5 using Jupyter Notebook version 6.5.4.