System Resilience Assessment and Improvement with Applications of Unmanned Aerial Vehicles
Natural and human-caused events often disrupt physical networks such as transportation networks, supply chain networks, and power networks. Natural disasters, political issues, economic crises, and even criminal activities can pose severe threats to the performance, endurance, and security of such networks. In order to enhance endurance and keep the networks' performance at a high level, it is necessary to increase and maintain the resiliency of these networks. Resiliency is defined as the ability of the network to respond to a disruptive event. Enhanced approaches to network surveillance can be of significant help for improving the security of such networks. Drone or Unmanned Aerial Vehicle (UAV) is a type of aircraft without a human pilot aboard. Accommodating cameras, sensors, and other information-gathering equipment allows the drones to provide high-quality information about the area under surveillance. Hence, they can play a significant role in monitoring the physical networks. In the context of network resiliency under disruption, various definitions exist in the literature that consider different resiliency aspects. The first contribution of my dissertation research proposes a quantitative approach for measuring the resiliency of network components as well as the network itself. The component resiliency is defined as a function of criticality, disruption frequency, disruption impact, and recovery time. The component resiliency reflects the effect of a component disruption on the network, and, the network resiliency is measured by the resiliency of the weakest component in the network. Therefore, enhancing the network resiliency is possible through reinforcing the weakest components in the network in sequence. Efforts to improve network resiliency often require financial resources. As such, an optimization model is further introduced to maximize the network resiliency under the budget constraint. The proposed approach can help decision-makers assess and compare the network resiliency status, identify and improve the weakest component in terms of resiliency, evaluate the cost of resiliency improvement, and determine how much improvement can be achieved under a given budget limitation. If the nation's border is considered as an example of a physical network under threat, then providing continuous surveillance over the border can enhance its security. Thus, the second contribution is to propose a risk-based surveillance model using drones for border patrol. Using drones can help patrol areas that are typically inaccessible by field agents, reduce response time, and enhance the safety of the dangerous regions. However, the drones' short flight duration is a significant drawback for a seamless surveillance mission on the U.S. border. To address this, we propose using the dynamic wireless charging system equipped with the electrification line (E-line) to increase the drones' flight time in real-time. Hence, this research aims to optimize the length and location of the E-line in addition to the number of drones to cover the region of interest. By dividing the borderline into small segments, we consider each of these segments' risk and priority, and provide the decision-makers a set of candidate solutions for them to choose based on their preference and budgetary policies. A cost analysis method is then developed to propose a technique for decision-makers to decide among these candidate solutions. Power networks are critical infrastructures that should be under continuous inspection. Disruptions on power networks can have a severe impact on telecommunication, transportation, and energy networks. Continuous surveillance will help the technicians to be more quickly informed about the damaged areas and the details of the disruption. Furthermore, Using drones to survey the network can provide high-quality information about the health of the network. Moreover, the disrupted environment can be dangerous, and using drones can ensure the safety of the inspection workers. Therefore, the third contribution of this thesis proposes a drone routing framework to enable the systematic and automatic assessment of power networks considering wireless charging of the drone during the scanning. This chapter incorporates the resilience-oriented line priority index to periodically prioritize the power lines based on their specifications and conditions, which in turn improves the efficiency of the assessment.