Conventionally, community energy management system (CEMS) is provided with the information of surplus and shortage amounts only at each time interval. This limited information may lead to an increase in the operational cost of the multimicrogrid (MMG) systems. This paper suggests informing the CEMS about the adjustable power also, in addition to the surplus and shortage information. This additional information will result in a variety of options for the CEMS to fulfill the load demands of its network. CEMS will choose among various available options, which include trading with the power grid, buying from a controllable distributed generation plant, buying from a community battery energy storage system (CBESS), or controlling the adjustable power: increasing or decreasing the generation of controllable units. CBESS can either be controlled by CEMS or can act as an autonomous entity. The effects of both the operational options have been analyzed and economically efficient mode is suggested for MMG systems. Demand response (DR) is also considered in the proposed model. The incorporation of DR will ensure the supply reliability of the MMG system in addition to the reduction in operational cost. In contrast to the conventional single or two-step multimicrogrid optimization algorithms, a multistep hierarchical optimization algorithm based on a multiagent system is proposed in this paper. Easy to implement and computationally inexpensive mixed integer linear programming models are developed for each step.

Q-learning-based operation strategies are being recently applied for optimal operation of energy storage systems, where, a Q-table is used to store Q-values for all possible state-action pairs. However, Q-learning faces challenges when it comes to large state space problems, i.e., continuous state space problems or problems with environment uncertainties. In order to address the limitations of Q-learning, this paper proposes a distributed operation strategy using double deep Q-learning method. It is applied to managing the operation of a community battery energy storage system (CBESS) in a microgrid system. In contrast to Q-learning, the proposed operation strategy is capable of dealing with uncertainties in the system in both grid-connected and islanded modes. This is due to the utilization of a deep neural network as a function approximator to estimate the Q-values. Moreover, the proposed method can mitigate the overestimation that is the major drawback of the standard deep Q-learning. The proposed method trains the model faster by decoupling the selection and evaluation processes. Finally, the performance of the proposed double deep Q-learning-based operation method is evaluated by comparing its results with a centralized approach-based operation.

This paper presents an energy management strategy for day-ahead scheduling of networked microgrids and is named as nested energy management system. In the proposed nested energy management strategy, the surplus power existing in the inner level microgrids is reflected as a resource and deficit as a load to the outer level microgrids. The operation cost of networked microgrids with nested energy management systems is lesser as compared to conventional decentralized or hybrid energy management strategies. Similarly, due to the inclusion of data (surplus/deficit) from inner level microgrids, customer privacy also increases. In islanded mode, the resilient performance of disconnected microgrids is enhanced due to their ability to form sub-groups. The probability of subgrouping for L-layers of microgrids is given by (L-2)/L. In grid-connected mode, various cases with diverse generation costs at each microgrid level along with the time-of-use market price signals have been evaluated. In islanded mode, the effect of battery energy storage system and disconnection of microgrids from the network have been evaluated. Hybrid ac/dc microgrids have also been considered to evaluate the effectiveness of the proposed approach. Simulation results have proved the viability of the proposed approach.

In order to determine the optimal size of energy storage system (ESS) in a fast electric vehicle (EV) charging sta-tion, minimization of ESS cost, enhancement of EVs’ resilience, and reduction of peak load are considered in this study. Especially, the resilience aspect of the EVs is focused due to its significance for EVs during power outages. Firstly, stochastic load of the fast charging station (FCS) and the resilience load of the EVs are es-timated using probability distribution functions. This information is utilized to maintain energy level in the ESS to ensure resilience of EVs during power outages. Then, the annualized cost of the ESS is determined using the annual interest rate and life-time of ESS components. Finally, the optimal ESS size is determined using the annualized ESS cost, penalty cost for buying power during peak hours, and penalty cost for resilience violations. Simulations along with sensitivity analysis of uncertainties (market price, arrival time of EVs, and residual energy level of EVs), number of EVs in the FCS, and converter ratings are conducted. Simulation results have shown that increasing the penalty cost for peak intervals is a viable solution to decrease the peak load while controlling the cost of the FCS.

A resilience-oriented optimization strategy is proposed in this paper by considering feasible islanding in normal operation and survivability of critical loads during emergency period. A resilience index is proposed for assessing the capability of each microgrid to feed local critical loads during sudden power disruptions. Based on the values of the proposed resilience index, local optimization results are revised by changing the commitment status of dispatchable generators and energy storage elements. The uncertainties associated with renewable generations and loads are realized via robust optimization method. The adjustable power bounds, suggested by each microgrid for minimizing the operation cost of the network, are also revised through the proposed resilience index. In emergency mode, decision between feeding of lesser critical loads and battery charging along with a strategy for minimization of load curtailment during switching of scheduling windows is considered. These two considerations assure the survivability of critical loads during emergency period. Finally, an incremental cost consensus algorithm is used for optimal allocation of surplus power among the connected microgrids having unserved loads. Three different network topologies are considered in emergency mode for assessing the performance of the proposed approach.