Non-Intrusive Load Monitoring (NILM) refers to as the technology of identifying the operation status and power consumption of individual electrical devices (typically household appliances) from an aggregated smart meter reading profile. This paper proposes a novel multi-appliance adaptive NILM framework designed for simultaneous disaggregation of multiple appliances with high memory and computational efficiency. By developing a self-adaptive loss weighting strategy, the proposed technique adaptively balances the contributions of regression and classification losses across appliances, preventing performance biases toward high-power or frequently operating devices. The NILM system integrates an encoder-decoder architecture and attention mechanisms to extract appliance-specific features while ensuring robust generalization. Additionally, the framework supports a self-supervised mode as an alternative deployment strategy, allowing the model to train a generic representation using only aggregated power data, reducing dependence on fully labeled datasets. Evaluation on the REDD and UKDALE datasets demonstrates improvements in both estimating power consumption and detecting appliance states.

Passive daytime radiative coolers (PDRCs) are an effective solution for mitigating urban overheating and minimising building cooling energy consumption. Phase change material-enhanced radiative cooling (PCM-RC) roofs can further enhance summer passive cooling and mitigate winter overcooling. Most prior studies on PCM-RC roofs rely on short-term, summer-focused, material-level experiments for proof-of-concept demonstrations, leaving their year-round performance inadequately evaluated. Therefore, a reliable roof model is critically needed for long-term performance assessment. However, existing simulation models commonly neglect surface condensation and evaporation, even though these phenomena influence latent heat transfer and thermal performance. Here, a coupled thermal-dew model for year-round PCM-RC roof simulations was developed and validated against outdoor measurements. The discrepancies during evaporation and condensation accounted for a maximum 4.9% and 8.1% of the daily heat transfer and were further analysed to clarify the model limitations. The model was used to perform annual simulations of four representative roof types in Sydney’s climate, incorporating PDRC, PCM-RC, and reference coatings. Compared with the conventional model (without surface condensation/evaporation), the thermal-dew model predicted up to a 10% increase in the annual through-roof heat loss for PDRC roofs and a 4% increase for PCM-RC roofs. PCM-RC roofs exhibited up to 20% and 49% lower annual through-roof heat transfer than PDRC and reference black roofs, respectively, while maintaining comparable urban overheating mitigation performance. This work presents an experimentally validated thermal-dew model for year-round evaluation of PDRC and PCM-RC roofs, applicable to building performance simulation and dew collection analysis of radiative coolers.

This paper presents an experimental study aimed at evaluating the hygroscopic behavior of wood for potential application in adaptive façade shading systems. Due to its intrinsic ability to absorb and release moisture, wood offers promising prospects for biomimetic designs inspired by the actuation mechanisms observed in conifer cones. A series of five experiments were conducted in a controlled climatic chamber, systematically varying humidity between 30

Power load super-resolution technology aims to recover power load time series data with a high resolution from low-resolution power load data, providing a more accurate representation of actual power load. Existing deep learning-based load super-resolution methods require paired low-and high-resolution load data to train a model, and the model is used to perform load super-resolution tasks on the unseen low-resolution load data. Nevertheless, these approaches have limited applicability in real-world scenarios where high-resolution data is difficult to obtain, which makes training the model infeasible. Motivated by this, this paper proposes a new system to overcome this limitation. Supported by self-learning techniques, the system can generate load data with a desired high resolution without using high-resolution load data to train the load super-resolution model (i.e., the model can only be trained by the low-resolution load data). The system is also equipped with an unsupervised noise identification module, enabling the system to effectively identify diverse noise signals in the load sampling process and improve the reconstruction accuracy. Experiments and comparison studies are conducted on a real-world dataset to validate the effectiveness of the proposed technology, and the results demonstrate the technology can achieve a comparable performance to supervised load super-resolution methods and has high adaptivity to load super-resolution tasks with different scaling factors.

Urban surfaces absorb heat, raising ambient temperature and contributing to urban overheating, greenhouse gas emissions, environmental pollution, energy hikes, and heat-related health issues. Passive Daytime Radiative Coolers (PDRC), available in white or silver colours, are characterised by very high solar reflectance and emissivity in the atmospheric window and offer sub-ambient cooling. Nevertheless, optical annoyance and aesthetic issues limit their application to high-rise buildings. Winter overcooling induced by the highly reflected PDRC application is another concern, mainly in heating-dominated climates. To address these issues, an alternative approach dubbed fluorescent-based Passive Coloured Radiative Coolers (PCRC) offers an additional fluorescence-based cooling mechanism. By absorbing solar radiation and then reemitting light, they provide colour, avoiding absorbing pigments and reducing thermal balance and visual discomfort. Three fluorescent- based prototype PCRCs in orange, green and red colours were developed using quantum dots and fluorescent dyes and tested under the hot desert climatic conditions of Alice Springs (Australia) in the autumn and the humid climatic conditions of Sydney (Australia) in the winter to assess their year-round thermal performance. Overall, the best cooling performance was observed for the orange-fluorescent prototype PCRC. Under dry-hot conditions in Alice Springs, this prototype achieved the same surface temperature as highly reflective corresponding PDRCs and 2.7 degrees C higher cooling than the conventional white cooling roofing membrane during the daytime. During the night, all PCRCs were 7-8 degrees C below the ambient temperature. Under the humid-cold Sydney's winter conditions, the orange-fluorescent film yielded improved insulation, roughly 4.5 degrees C higher than the white reference. Under cold conditions, higher PCRC surface temperatures were attributed to the lower IR transmittance of polymers in which fluorescent dyes or quantum dots were embedded. These results are encouraging, as implementing such PCRCs may support achieving a balanced performance during both summer and winter and expanding the use of radiative coolers in urban environments by addressing optical annoyance and aesthetic issues.

The proliferation of decentralized applications across different autonomous blockchains raises the need to enable cross-chain data interoperability (CCDI). However, prior approaches for supporting CCDI often hit scalability bottlenecks regarding critical metrics, e.g., memory, or remain prone to withholding and censorship attacks. This paper proposes two protocols to implement secure and efficient CCDI under adversarial conditions. The cross-chain token exchange (CCTE) protocol for atomic swaps is proposed. It adopts a deposit mechanism, a blockchain-of-blockchains (BoB), and Merkle proofs to ensure the completion of token exchanges even under withholding attacks. It utilizes a parallelized design to support concurrent token exchanges, thereby improving its efficiency and avoiding censorship attacks that target sequential token exchanges. The CCDI protocol is proposed to support any CCDI application. It authorizes a unique BoB to execute arbitrary CCDI application logic. It integrates a “transfer and in place data update” mechanism to improve its efficiency, and this mechanism enables a blockchain update its state data items using a single transaction, without requiring any information from other blockchains. Moreover, the CCDI protocol integrates a state data migration scheme, which supports a user to migrate its state data item to censorship-resilient blockchains, and incorporates a malicious user nodes elimination scheme, which enables the updates of state data items in a CCDI process even under withholding attacks. Systematic performance evaluations are conducted to compare the two protocols with existing ones. The CCTE protocol reduces latency by at least 52% compared to existing protocols under probabilistic consensus setting. The CCDI protocol outperforms prior protocols, lowering communication cost by 59%, computation overhead by 41%, memory burden by 12%, and latency cost by 33%.

Non-Intrusive Load Monitoring (NILM) is a technology that identifies the operational status and power consumption of individual electrical devices (typically household appliances) from aggregated smart meter readings. This paper introduces a self-adaptive loss weighting strategy aimed at improving the performance of a single deep neural network-based NILM model for disaggregating multiple appliances. The proposed technique adaptively balances the contributions of each appliance, preventing performance biases toward high-power or frequently operating devices. Additionally, we design and implement a multi-appliance adaptive Autoencoder framework to evaluate this mechanism. Experimental results on the REDD and UKDALE datasets show improvements in disaggregating the power consumption of multiple appliances.

Non-Intrusive Load Monitoring (NILM) refers to as the technology of identifying the operation status and power consumption of individual electrical devices (typically household appliances) from an aggregated smart meter reading profile. This paper proposes a generative, transferrable NILM system called “GT-NILM”, which can provide accurate NILM services to stakeholders (e.g., grid operators and building managers) for understanding the operational information of individual appliances of residential households while sufficiently preserve the households energy data privacy. The proposed system is backboned a conditional diffusion model and a convolutional neural network, which are designed for identifying the appliances operating power waveform characteristics and ON/OFF status from smart meter readings, respectively. This dual-model design enables fast model training while maintaining high NILM accuracy, with performance validated on REDD and UK-DALE datasets. In standard within-dataset evaluations, GT-NILM achieves comparable performance to state-of-the-art NILM models, while in a cross-dataset transfer learning experiment, it outperforms other state-of-the-art methods up to 10.95% on the F1-score metric and 36.31% on the mean absolute error metric.

With the emergence of hydrogen production and refueling infrastructures for hydrogen fuel cell vehicles, planning and operation of Energy-Transportation-Integrated Hydrogen Networks (ETIHNs) starts to draw attention in academia. This paper proposes a comprehensive planning framework for ETIHNs equipped with Wind-Solar Hydrogen Production Stations (WSHPSs) and Hydrogen Refueling Stations (HRSs) that deliver hydrogen by tanker vehicles. Models of the critical components in a ETIHN are established, and a dynamic hydrogen logistics system model is developed to enable reliable hydrogen transportation in both hydrogen supply and demand sides. Based on these models, a ETIHN planning model is formulated, which aims at minimizing the system's annual cost while satisfying the system's operational constraints. A CPLEX-based optimization approach is developed for solving the proposed planning model and numerical case studies are conducted to validate the proposed technique.

Over the last couple of decades, significant efforts have been made to develop structural health monitoring solutions. The growing need for the dynamic characterization of structures supports the implementation of condition assessments, maintenance, and monitoring strategies for existing and new civil engineering structures, and to provide increased safety for the public. Wireless monitoring systems are still being improved as the technology is finding a wider use for the monitoring of civil engineering structures, thanks to their easier installation and reduced costs when compared to the wired counterparts. In this context, this paper presents a new wireless network system for the dynamic characterization of civil engineering structures, whose distinguishing features comprise combining cutting-edge accelerometers, excellent signal synchronization, low battery consumption nodes, and a cloud-based framework to support the monitoring operations. The performance characteristics are validated through laboratory tests and are demonstrated on a newly constructed 211 m tall building.

Passive Daytime Radiative Coolers (PDRCs) can cool buildings and cities by achieving subambient surface temperatures thanks to high solar reflectance and selectively high infrared emissivity within the atmospheric window. However, their performance is assessed outdoors with often too small samples and inconsistent experimental setups across the literature, which hinders intercomparison. Here, we demonstrate the limitations of current approaches and identify an experimental setup to achieve consistent and comparable results. We measured the surface temperatures of PDRC samples with different setups under both above-ambient and subambient conditions in temperate and desert climates, focusing on the effects of sample size and polyethylene (PE) film cover. Sample size introduced significant measurement uncertainty for air-exposed PDRC, resulting in overestimations of 1.2°C and 1.6°C under above-ambient and sub-ambient conditions, respectively. Further, PE film covers introduced greater deviations, up to + 3.0°C for above-ambient and − 2.7°C for subambient PDRCs. Instead, air-exposed PDRC samples of at least 200 mm, placed over insulation boards with reflectivity comparable to the PDRC, and with surrounding buffering boards, achieve a low measurement uncertainty of ± 0.3°C. Compared to this stable setup, other setups exhibit a large measurement bias, with 90% of this bias being systematic. Our findings underscore the need for consistent outdoor measurement methods to assess PDRC performance and enable quantitative analysis across the literature.

This chapter introduces the concept and key features of a building energy management system (BEMS). Particular attention is provided at describing how a BEMS interacts with energy resources deployed in a building and with renewable energy resources. The most common categories of building-site energy resources are outlined to highlight their key characteristics. An application example is then considered to highlight a possible design strategy of a BEMS to monitor and control energy storage capabilities available at the building level and contributions from renewable energy resources. The relevant model representations for each of these resources are presented to enable simple operational algorithms to be implemented.

With increasingly prevalence of distributed renewable energy sources, Peer-to-Peer (P2P) energy trading has become an active research direction. This study explores the role of the participants' Socio-Demographic Characteristics (SDCs) in the decision-making process P2P energy trading by proposing a personalized P2P energy trading system. The system periodically collects the participants' bids and pair energy sellers and buyers to form transactions. An attention-based SDC inference system is developed, which identifies a participant's SDCs from the on-site historical smart meter readings. Followed by this, the system analyzes the importance of energy buyers' demands based on their SDCs, and an alternative current network constrained P2P energy market clearing model is formulated to maximize the participant population's social warfare by considering their energy demand importance and economic benefits. Simulations based on real-world datasets are conducted to validate the proposed system.

Along with the widespread deployment of distributed energy resources, Peer-to-Peer (P2P) energy trading has become an active research topic. Driven by the fact that in real world, people's willingness on P2P energy trading would be affected by multi-fold factors (including both financial and non-financial factors), this paper proposes a personalized P2P energy trading system that facilitates energy trading among the participants by sufficiently considering their energy trading profit/cost, social relationships, and personal features. A Graph Convolutional Network (GCN)-based network analysis model is utilized to infer the matching degrees between two participants; based on this, an auction-based P2P energy market clearing model is proposed to maximize the participant population's social welfare. AC network constraints of the underlying grid are incorporated into the market clearing model to ensure the physical feasibility of the energy trading transactions and security of the grid; this makes the system applicable to different scales of P2P energy trading (e.g., citywide and local community scales). Numerical simulation is conducted based on the IEEE 33-bus distribution system to validate the effectiveness of the proposed system.

This chapter presents the information infrastructure that supports the operations of building energy management systems in buildings. In the first part of the chapter, building automation systems (BASs) are introduced, and their components are briefly presented to outline how these can support the operations and strategies of building energy management systems (BEMS). This is followed by the introduction of the concept of Building Internet-of-Things within the Internet-of-Things paradigm and how this can assist the implementation of effective BEMSs' strategies. In the second part of the chapter, cloud and edge computing are described to highlight how they can optimize the computations required by a BEMS. The final part of the chapter is dedicated to the information infrastructure that combines the paradigms of BAS/BIoT, cloud computing, and edge computing.

This chapter intends to provide a brief overview of the energy sources that are typically deployed in buildings and that will be considered in coming chapters when dealing with the design of building energy management systems. In the first part of the chapter, we present key features related to wind power and outline common typologies of wind turbines together with design considerations related to their implementation in urban environments. This is followed by an overview of the solar energy and how this can be harvested through different technologies available for building applications. In the final part of the chapter, we present key characteristics of energy storage systems that have been gaining popularity in recent years with the wider penetration of renewable energy sources.