XXX-X-XXXX-XXXX-X/XX/$XX.00 ©20XX IEEE Development of Open-Source, Edge Energy Management System for Tactical Power Networks Syed Wali Department of Electrical and Computer Engineering Texas A&M University College Station, TX, USA syedwali@tamu.edu Muhammad Areeb Fasih Department of Electrical Engineering NED University of Engineering and Technology Karachi, Pakistan areebfasih84@gmail.com Irfan Khan Department of Electrical and Computer Engineering Texas A&M University College Station, TX, USA irfankhan@tamu.edu Muhammad Hassan ul Haq Department of Electrical Engineering NED University of Engineering and Technology Karachi, Pakistan hassanulhaq@neduet.edu.pk Yasir Ali Farrukh Department of Electrical and Computer Engineering Texas A&M University College Station, TX, USA yasir.ali@tamu.edu Majida Kazmi Department of Electrical Engineering NED University of Engineering and Technology Karachi, Pakistan majidakazmi@neduet.edu.pk Abstract— Microgrids specialized for tactical operations have been subjected to several challenges. These tactical power networks are islanded and have a relatively low power generation capacity. Meeting power requirements of military equipment, having intermittent and highly inductive nature, exposes microgrids to severe stresses. Existing methodologies to monitor and control the impact of load variations require sophisticated equipment and trained personnel. The objective of this research paper is to present an open-source edge energy monitoring system (EEMS) for efficient demand management of tactical networks. The proposed system is capable of capturing all minute operational artifacts, including harmonic distortions and power quality of these networks. A variable gain amplifier circuit enables the proposed EMS to sense all the signals in a wide range of power with higher resolution. The proposed system utilizes raspberry pi as an edge device to meet the low power requirements of tactical networks. The novel concurrent programming approach adopted in the proposed EMS, effectively handles the large amount of data acquired from the network. This parallel processing of acquired data speeds up the execution process. All electrical parameters obtained during this process are stored in an encrypted local database that can be utilized for fault analysis and load prediction. Further integration of machine learning tools in proposed EMS assists in automated power network reconfiguration and tuning under harsh battlefield situations. Keywords— Edge computation, Tactical Networks, Data Management, Parallel Processing, Electrical Parameter. I. INTRODUCTION Reliable power is crucial for the operational resiliency of command-and-control centers. There are two different strategies to distribute power in tactical networks. One strategy is known as spot generation, where a single power source directly feeds military loads. Another common strategy is regarded as microgrids, where several synchronized generating resources meet the power requirements of military loads via a distribution network. Fig. 1 represents a typical microgrid for tactical networks where all power sources are connected to the electrical feeder for serving the load demand of military equipment. The parallel arrangement of generating units in microgrids ensures continuity of power under any contingency. They allow maintenance on one or more generators while other resources fulfill the power demand of the network. However, these microgrids create unique challenges in the military environment. These islanded power systems do not have large power generating units that can behave as system regulators under harsh battlefield situations. Generator sets (Gensets) utilized in this network usually vary in the range of 30kW to 60kW [1], whereas the connected load is highly inductive and intermittent in nature. Therefore, whenever a sudden load change occurs on the microgrid, large power oscillation takes place between synchronized generating units, failing the complete power network. This issue occurs due to the synchronization of generating units from different manufacturers in tactical networks [1]. One of the possible solutions to cater to this problem is an immediate response from the power operator. However, it requires highly trained personnel and sophisticated equipment of proprietary nature [2]. Moreover, different types of proprietary equipment are also incompatible with each other. They cannot fetch data from all types of equipment used in the system. Therefore, no adequate performance history of equipment is maintained in these networks. As a result, performing manual root-cause analysis is error-prone, even experienced personnel may take several hours in achieving this objective. The only possible solution for resolving problems of the tactical power network is to develop an energy monitoring system specialized for tactical operations. This system should monitor all artifacts of the electrical power network without being dependent on proprietary equipment. Furthermore, since tactical networks do not have abundant power resources to energize heavy devices, these networks must be monitored using low power-consuming edge devices. Therefore, this research proposed an edge energy management system (EEMS), with the main contributions as follows: • Open-Source: The proposed EEMS for tactical networks is based on open-source hardware and software platforms that eliminate the dependency on proprietary equipment. As the proposed system is not vendor specific, it is compatible with any type of tactical network. • High-Resolution: Data acquisition system of the proposed EEMS contains a variable gain amplification circuit, capable of capturing current and voltage phasors with high resolution. Processing of this discrete-time discrete-valued (DTDV) signals gives a comprehensive knowledge of the power network. It can also be utilized for training machine learning predictors for diagnosing the performance issues of the tactical network.