Temporal series forecasting is challenging due to the lack of persistence of excitation conditions imposed by William's lemma, which makes the system identification task very difficult. This work presents the delay embedded dynamic mode decomposition (DeDMD) technique to eliminate the persistence of excitation problems and to make accurate forecasting even with limited training data set. The existing DMD algorithm is enhanced to make it more relevant for the data which is not persistently excited by artificially expanding the dimensions of spanned subspace through embedding with Hankel matrix. The methodology is tested in a case study for forecasting the power injected by distributed energy resources (DERs) in the microgrid. Stability analysis for different Hankel rank conditions in addition to error analysis for both training and testing data is also performed to support the adequacy of DeDMD. The observed results suggest that the proposed method gives results with minimum error for temporal series prediction. The critical contribution of the paper is to extend the DMD technique's applicability for the data that is not persistently excited.
Integration of large-scale distributed generators (DGs) (such as wind turbines, solar photo-voltaic, fuel cells, and so on) into the microgrid triggers voltage deviation which can give rise to desynchronization and distortion of power flow. To solve this problem, this article presents a distributed optimal secondary voltage control to synchronize the output voltage deviations of the DGs to zero. Instead of designing a distributed linear optimal control for the DGs linearized near an operating point as in the existing literature, a two-loop control structure comprising an inner-loop and outer loop is put in place. The inner-loop linearizes the nonlinear dynamics of each of the DGs. Then, the outer-loop which consists of the optimal distributed controller solves an optimization problem with the aim to minimize the voltage deviation of each DG by exchanging information with neighbouring DGs through a communication network. The proposed control technique will force the voltage deviations of the DGs to zero based on consensus. The efficacy of the proposed technique is confirmed by simulation studies.
Autonomous robots' applications have been increasing in various fields, including healthcare. However, the use of autonomous robots has posed various challenges in terms of cost, navigation and operational complexity. Many researchers have explored the applications of autonomous robots in healthcare services. Some developed virtual assistants, like Chatbots, Hospital Care Watch (HCW), Intelligent Diabetes Assistant (IDA), etc. Others focused on developing robot navigation systems. However, those research works have not focused on developing healthcare assistants that would accompany patients while visiting medical facilities. This paper presented one of the vital applications of autonomous robots in healthcare services. The primary aim of the work is to develop a low-cost autonomous healthcare Digital Assistant Lead RoBot (DALBOT) to serve patients and guide them to the desired location while visiting medical facilities. The locations were identified using codes with corresponding Cartesian coordinates. The results demonstrated the feasibility of integrating DALBOT with other healthcare systems, such as stretchers, wheelchairs, medical supply deliveries, etc.
With the development of the Internet of Things (IoT) industry, many services and applications have been developed, such as indoor localization, proximity detection, smart home, and so on. However, many services are severely limited due to energy constraints. Many researchers have proposed solutions to reduce energy consumption in these applications and improve energy utilization. For instance, Bluetooth low energy (BLE)-based data transmission scheme and sensor-based energy management scheme. In this paper, we combine BLE and sensors to propose an indoor localization-based energy management system for smart homes, which can effectively solve the energy waste caused by users forgetting to turn off the device's power. The analysis results show that the proposed scheme can save energy effectively.
A joint scheduling model of load-side voyage and power generation is proposed in this paper in order to realize energy efficiency and emission reduction for all-electric ships. The optimization model contains several nonlinear constraints, such as propulsion power fluctuations caused by sailing speed variations and equipment lifespan losses caused by energy storage dispatch. The piecewise linearization method is an effective way to reduce model nonlinearity, but a large number of integer and binary variables introduced weakens the model convexity and increases the computational burden. Therefore, this study proposes an improved piecewise linearization method based on differential evolution to balance the number of decision variables and computational accuracy. In addition, the degradation loss caused by irregular charging and discharging of energy storage is quantified as the economic cost to optimize the energy storage operation strategy. Furthermore, the impact of greenhouse gas emission limitations on the power generation schedule is considered. Finally, the simulation of the proposed method is validated based on an all-electric ship.
Hydrogen electrolyser loads pose to add tremendous demand to the Australian National Electricity Market (NEM) given the accelerating energy transition and opportunity for widespread decarbonisation. Accordingly, a consideration for flexible integration of future electrolyser loads is paramount while ensuring the energy consumed is considered `green'. This paper adopts a linear optimisation model to counterfactually investigate flexible participation strategies of these loads considering power purchase agreement (PPA) structures and the large-scale generation certificate (LGC) scheme in a case study utilising historical NEM data. It demonstrates both challenges and opportunities for flexible load operation given the ability to harness price lulls, but similarly be exposed through unmatched hedges given variable volume PPAs. Since the ability to maximise hydrogen production is limited in this study to the traded green-certificates under an assumed bundled PPA structure, this paper reveals benefit in pursuing dual PPAs with both solar and wind generators. Finally, the study yields a key question as to the certification of green-hydrogen and further need for methods to quantify temporal matching of generation and consumption.
This paper presents a real-time, coordinated approach for residential demand response (DR) participation in electricity markets under the dynamic operating envelopes (DOEs) framework. In the first stage, the distribution network service provider (DNSP) utilises a convex hull method to construct DOEs at each customer point-of-connection (PoC). In the second stage, the demand response aggregator (DRA) employs a hierarchical control framework for tracking a load set-point signal commanded by the market operator while individual households minimise their electricity costs. In this regard, the real-time control operation takes account of DOEs assigned by the DNSP. The simulation results validated on a real Australian network with realistic data suggest that the DRA is able to achieve precise tracking of the load set-point signal while honouring network statutory limits and also managing comfort for end-users. Furthermore, the approach preserves end-user privacy and is scalable.
Transmission and distribution networks based on medium voltage direct current (MVDC) solutions have the potential to provide higher power quality, greater power transmission/conversion efficiency, enhanced power supply reliability/stability, improved power supply capacity, and more efficient corridor utilisation compared to an equivalent ac system. Moreover, MVDC networks can facilitate the integration of renewable energy sources (RESs), energy storage systems (ESSs), and the increasing number of dc loads. However, these systems are still in early development stages; their advantages in need of clear demonstration before large-scale commercial adoption. This paper explores general MVDC network structures and different application scenarios focusing on cases of particular interest to Australian power systems and to Australia in general. Areas of interest are identified and equipment requirement, protection schemes and related challenges are also discussed, highlighting both technical barriers and possible development directions.
As the power system is undergoing fundamental changes in the scope of the energy transition, the existing regulatory structure lacks the necessary components to cope with these changes. More decentralized energy resources are being integrated into the system on lower voltage levels, while coordination of these, especially considering the limited grid capacities, is not realized. Furthermore, central power exchanges operate without considering these physical capacities within one bidding zone and limit small-scale participants. A restructured energy market is required to enable small-scale asset integration and the recognition of physical limitations. Therefore, this paper presents a decentralized, grid-aware peer-to-peer market model that allows for automatically executed, local schedule procurement of peers within a distribution system community. The model uses nodal trading agents and a sensitivity-based approximation to assess the dynamic grid fees of any transaction, hence mapping the physical limitations to the consumers' energy prices. Therefore, users can fully adjust their behavior based on the demand and availability of energy and the grid capacities, which the energy prices comprise. Based on these, each peer is enabled to adjust its behavior through load shifting, storage optimization, and various other adaptions. Overall the proposed market structure transforms the power system from a correcting to a coordinating nature.
Some of the most challenging parts of finding new approaches to lowering residential energy use involve studying, detecting, and visualizing households' anomalous power usage patterns. This research presents a novel method for identifying irregularities in electric vehicle energy use by extracting features using a modified long-short-term memory model. The latter is implemented to extract load features by analysing intent-driven user consumption instances occurring throughout the day. In addition to feature extraction, we explore the use of a deep neural network, specifically the LSTM architecture, to efficiently detect and classify anomalies in Electric Vehicles. In the following, we provide a novel anomaly visualisation technique based on a scatter representation of the classes, which gives customers a simple way to comprehend unusual actions. These encouraging findings validate the effectiveness of the suggested deep learning approach for identifying abnormal energy usage, encouraging energy-efficient behaviour, and cutting down on energy waste.
Chaotic systems are nonlinear deterministic systems that reveal complex and unpredictable behavior. Chaos theory has been exploited for analysis and synthesis in several academic and technical domains, including communication, economic systems, electrical systems, chemical processes, and optimization. A crucial step in analyzing the behavior of a complex system is identifying the chaotic system's parameters. The problem of parameter estimation of the chaotic system is identified and a solution via finite-time estimator has been proposed without the hypothesis that the regressor is persistently excited. As a finite-time estimator, the I-DREM approach is explored to estimate the parameters of uncertain chaotic systems. This paper analyzes the chaotic systems parameter estimation problem under both PE and non-PE conditions for comparison purposes. Additionally, the Lorenz system's unknown parameters are all identified. Finally, numerical simulation results are displayed to show exactly how well the proposed strategy performs.
Mobile power sources (MPS), such as electric vehicles (EVs), potentially improve distribution network (DN) restoration under extreme event conditions. However, employing EVs as a major power source is under researched, as is the prepositioning, routing, and dispatch of large numbers of EVs. This study proposes a three-stage optimization approach to achieve proactive prepositioning, dynamic routing, and dynamic power scheduling for effective assessment of resilience and needy restoration. First, EVs are prepositioned in the DN to enable swift pre-restoration and improve the survival of loads. Second, following an extreme event, EVs are dynamically routed in the DN and transportation system (TS) to improve system recovery. This stage also proposes a novel EV travelling model, bridging consumption rate and distance to study the efficacy of EV's state of charge (SOC) and the participation decision of the EV's user. Third, dynamic power scheduling of EVs is addressed, based on decisions made in the previous two stages. A mixed-integer programming model that addresses matters such as various timeframes of EV dispatch and DS operation, and the connection of road and power networks, is tested via case studies of a three-phase AC IEEE 123-node test system to demonstrate the effectiveness of the proposal.
In this paper, a cell level balancing technique is presented for aqueous sodium batteries. The balancer consists of a series combination of a diode and a switch that is connected in parallel with each cell in a series connected string. The switches are used in series with the diodes to perform the balancing in the lower cell voltage range. This enables to account for the diode forward voltage variations due to several factors including temperature, manufacturing limitations, knee effect and parametric variations over the time. The variation in diode forward voltage drop could result in improper balancing specially at higher state-of-charge. As the balancing is only performed in the lower cell voltage range, it may take several charge/discharge cycles to balance the cells but at the end series connected cells get properly balanced. The control algorithm of the proposed diode based balancer requires only two voltage measurements i.e., battery voltage and voltage of the lowest cell in the series connected string. The simulation results are presented for the balancing of eight series connected cells in a string. The simulation results are promising and verify the effectiveness of the proposed balancing technique.
In 2021, the largest solar farm and renewable energy microgrid to date at an Australian university commenced operation. Located at the Waurn Ponds campus of Deakin University, the microgrid will be a key enabler for the University to achieve its carbon neutrality and 100% renewable electricity targets by 2025. The 14-hectare solar farm is capable of producing up to 7.25MW power which translates to 54% of the electricity demand of the campus. To enable real time system monitoring, data analysis, power output prediction, and simulation of the microgrid for answering what-if questions, a digital twin was developed in-house. In this paper, we present the Deakin Microgrid Digital Twin $(\mathbf{D}\mu \mathbf{DT})$ with a focus on the health monitoring functionalities (essentially a microgrid assessment tool), including: efficiency monitoring, alarm analytics, and Artificial Intelligence (AI)-driven anomaly detection. The design of the digital twin is general in nature, with software, machine learning and visualisation components designed to make it possible for utility extensions beyond the Microgrid application.
A driven-data trajectory approach has been developed to allocate dynamic VAr source (DVS) to improve the short-term voltage stability (STVS) of power grids. The siting approach for DVS would be carried out by the comparing grid responses of different sites with DVS by considering the desired reference response. The undergoing assessment emphatically covers the full signature of grid dynamics interaction involving generation, transmission, and load characteristics. For illustration, the developed approach is applied to the Reliability and Voltage Stability (RVS) test system designed for STVS analysis. Several scenarios are tested, such as different levels of induction motor load, large-scale PV (LSPV), and LSPV reactive current injection, to demonstrate the viability and robustness of the approach. Subsequently, the viability and robustness of the siting approach are verified by checking STVS performance using the VRIsys index.
For payment settlement, a blockchain-based Peer-to-Peer (P2P) energy trading requires a stable medium of exchange with little price volatility. Stablecoins, the most suitable medium of exchange, gaining concentration even from central banks. A consortium of central banks recommends complying with capital and liquidity standards for the high-quality liquid asset (HQLA) for the solvency of banks or financial institutions. Stablecoin as HQLA requires the adaption of such standards in P2P energy trading for reserve resilience. We propose a mechanism (NF90) that controls the inflow of stablecoins responding to Liquidity Coverage Ratio (LCR) for reserve resilience. Basel III accord recommends 100% of LCR. We measure the efficiency of NF90 concerning LCR as a metric. We simulate the proposed mechanism in Hyperledger Fabric as a permissioned blockchain platform for decentralisation, data storage, and smart contract. Experiments show that NF90 is the most efficient inflow control mechanism compared to other simulated mechanisms.
Microgrids (MGs) have gained popularity due to their many advantages in the integration of renewable energy sources to the power grid. With its capability to change its mode of operation, a microgrid can provide effective operation to supply electricity continuously to the customers. When the microgrid gets disconnected from the utility power system (islanding), it can exist as an independent entity by supplying the load with a sinusoidal voltage keeping its amplitude and frequency in the acceptable band by incorporating proper control. The islanding of microgrids may occur due to unpredicted natural calamities (unintentional) or by an intentional event. One of the key desired characteristics of a microgrid is its capability of achieving seamless transition from grid-connected mode (GCM) to islanded mode (ISM), and vice versa. It is necessary to supply power to the loads during the transition phase without losing its quality. Thus, it is mandatory to implement efficient control strategies to reduce the transients due to frequency and phase mismatches, which can lead to voltage and current variations during the transition phase. This paper presents a review on the literature related to the control schemes for seamless mode transition and how resynchronization to the utility can be established.
STATCOMs are an integral component for the stability of modern power systems; able to supply and sink reactive power on demand, they can assist with local voltage regulation. In this paper, a modulated finite control set model predictive control is developed and demonstrated, demonstrating voltage regulation through VAR control under different scenarios. The proposed modulated FCS-MPC is compared to a typical linear PI controlled STATCOM on a medium voltage distribution system, showing faster control dynamics and reference tracking, as well as decreased THD.
As residential battery energy storage systems (BESS) proliferate, it becomes increasingly important to be able to control them optimally within retail electricity market arrangements. Aggressively cycling the battery leads faster degradation and ultimately replacement, which is likely to be uneconomical. Accordingly, battery degradation should be considered in BESS control strategies. This paper uses a linear cost model based on total BESS throughput to incorporate degradation cost into the cost function of an optimal control strategy. This model is chosen due to its practicality in situations where in-depth battery parameters are unknown, real-time dispatch requires fast computation. It is shown that the inclusion of a degradation model has a significant impact on BESS operation under various residential retail tariff structures, suggesting that such models should be included in control strategies. Care is needed to ensure degradation costs are accurate since there is a high sensitivity in these strategies to cost assumptions.
