Possibilistic–Probabilistic Risk-Based Smart Energy Hub scheduling considering cyber security in advanced metering infrastructures

Future energy systems will be more efficient because of the energy transition of which, decarbonization, decentralization, and digitalization are important aspects. In addition, cyber security becomes more important in the digitalized energy system as cyber-attacks increase. Since the Smart Energy Hub (SEH) is well known as an efficient type of smart digitalized energy system, cyber security is of great importance in SEH operation. In this paper, a novel possibilistic–probabilistic​ SEH model is proposed to evaluate the effect of cyber-attacks on the demand side of the energy system. The presented model is called Possibilistic–Probabilistic Risk-based Smart Energy Hub scheduling model considering Cyber Security in the Advanced Metering Infrastructure (PPRSEHAMICS). It contains different uncertainties regarding renewable generations, demand response programs, gas fuel, wind turbines, and energy prices. The uncertainty of demand response programs is represented using the Z-number as a possibilistic–probabilistic method. Other uncertainties are represented using a scenario-based hybrid approach and Monte Carlo Simulation. Also, the downside risk constraint (DRC) approach is used to model the behavior of the SEH operator in the presence of uncertainties, so the operation and risk costs are simultaneously optimized. Furthermore, a novel Monte Carlo-based FDI detection/correction method is proposed in order to determine and mitigate the effects of cyber-attacks. The proposed model is formulated as a Mixed-Integer Linear Programming (MILP) problem and the GAMS environment is used to solve it. In this paper, different scenarios are considered to evaluate the effects of demand response (DR) programs, SEH operator behavior, false data injection, and correction on the operation cost and risk-in-cost of smart energy hub systems. In the first stage, the obtained results demonstrate that using DR accompanied by DRC yields the optimum operation strategy where the operation cost and risk are maintained at an acceptable point. Further, simulation results suggest that FDI attacks can deviate the operation cost and risk by a substantial amount, and the proposed Monte Carlo-based detection/correction method mitigates almost 79 percent of the false data while keeping the convergence speed in an acceptable range. Moreover, modeling the participation rate of the demand response program by the Z- number approach increases the validity of the results and the applicability of the proposed model.