7 July 2020 in IEEE Transactions on Industrial Informatics
Integration of power and gas systems has been recently proposed as a solution to deal with the sporadic availability of renewables and enhance the flexibility of power systems. In an integrated system, where critical operating information and control signals of both systems need to be communicated, the risk of cyber-attack is intensified. In this paper, we present a new model for integration of power and gas systems using power-to-gas (PtG) and gas-fired generation (GfG) facilities. We demonstrate how the operation of the integrated system can be impacted during undetectable cyber-attacks. We propose two new detection schemes for false data injection attacks against the input and output signals of the PtG/GfG facility scheduler. In the first scheme, a supervised machine-learning technique, based on convolutional neural network and wavelet transforms, is adopted to detect attacks on the information received by the facility scheduler. In the second scheme, a hybrid neural network is developed, based on an unsupervised learning technique, that requires no labeled training information to detect attacks on the output control signals issued by the scheduler. In both schemes, information acquired from local sensors and estimation methods is utilized for signal evaluation. The proposed schemes are incorporated into the facilities' scheduler to create a cyber-attack resilient scheduling model in an integrated power and gas grid. The efficacy and feasibility of the proposed model are evaluated via numerical studies using standard power system integrated with gas grid as the test bed using historical operating parameters.
11 March 2021 in IEEE Transactions on Industrial Informatics
A multistage approach to passive islanding detection is proposed that utilizes a decision tree (DT) like classification algorithm. The novelty of the proposed method is centered on the way in which features are passed to subsequent stages of the DT. Feature sets extracted using different sized time windows are passed to successive stages of the tree. This provides two important advantages: 1) cases that can be easily determined as either islanding or nonislanding events are flagged as soon as possible without waiting for the full feature set to become available; 2) because the algorithm allows for the use of different sized time windows, features are analyzed in time-scales that fit their natural patterns of temporal evolution. In this article, the proposed classifier is trained and tested using a database of feature vectors, obtained using PSCAD, which were designed to reflect a variety of commonly encountered events on an IEEE 34-bus distribution system. One of the key requirements for the proposed algorithm was that easy cases should be flagged as soon as possible; this property was confirmed by the observation that most events ( $\approx$ 79%) were detected within 10–20 ms, while at the same time retaining a very high detection rate overall cases ( $>\!99$ %).
30 April 2019 in GeoInformatica
19 June 2020 in IEEE Open Access Journal of Power and Energy
The recent legalization of cannabis is facilitating very rapid growth in the cannabis cultivation industry, with the energy intensive indoor cultivation facilities becoming more prevalent. This presents a challenge to utilities as the high energy demand from this industry can overburden the existing utility infrastructure. Hence, from both planning and operational perspectives, it is crucial to understand the energy consumption of the rapidly growing load. This paper proposes a deterministic energy consumption model for indoor cannabis cultivation operations for the two major loads in these facilities, i.e., lighting and HVAC, over a 24-hour period based on equipment specifications and operational requirements of the facility. This model can further be used to estimate or forecast short-term and long-term energy demands and costs of indoor cannabis operation(s). The proposed model successfully simulated the environmental conditions in a real-world cannabis facility, and the model’s energy consumption output is validated using actual measurements taken from this facility as well as model output using GridLab-D.
1 December 2019 in Electric Power Systems Research