Mario Berges

Enhancing Electricity Audits in Residential Buildings with Nonintrusive Load Monitoring

Nonintrusive load monitoring (NILM) is a technique for deducing the power consumption and operational schedule of individual loads in a building from measurements of the overall voltage and current feeding it, using information and communication technologies. In this article, we review the potential of this technology to enhance residential electricity audits. First, we review the currently commercially available whole-house and plug-level technology for residential electricity monitoring in the context of supporting audits. We then contrast this with NILM and show the advantages and disadvantages of the approach by discussing results from a prototype system installed in an apartment unit. Recommendations for improving the technology to allow detailed, continuous appliance-level auditing of residential buildings are provided, along with ideas for possible future work in the field.

Learning Systems for Electric Consumption of Buildings

Individual appliances’ electricity consumption is automatically disaggregated from a single custom metering system on the main feed to an occupied residential building. A data acquisition system samples voltage and current at 100 kHz, then calculates real and reactive power, harmonics, and other features at 20Hz. A probabilistic eventdetector using the generalized likelihood ratio (GLR) matches human-labeled events to the time-series of features. Machine-learning classification was most successful with the 1-nearest-neighbor algorithm, correctly identifying 90% of the laboratorygenerated training events and 79% of validation examples. The challenge of obtaining adequate training data for the real-world home leads to the development of the Wire Spy, a wirelessly-networked event detector with an inductive sensor which clamps to the cable of an appliance.

Contactless sensing of appliance state transitions through variations in electromagnetic fields

Non-Intrusive Load Monitoring (NILM) is a promising technique for disaggregating per-appliance energy consumption in buildings from aggregate voltage/current measurements. One major limitation of the approach is that it typically requires a training phase during which users must manually label device transitions. In this paper, we present an inexpensive contactless electromagnetic field (EMF) event-detector that can detect appliance state changes within close proximity based on magnetic and electric field fluctuations. Each detector wirelessly transmits state changes to a circuit-panel energy meter, which can then be used to label and disambiguate appliance transitions detected from the aggregate signals as well as to track the associated energy consumption. Our EMF sensors are able to detect significant power state changes from a few inches away making it possible to externally monitor in-wall wiring to devices (e.g., overhead lights). We experimentally evaluate our proposed EMF sensor in terms of power consumption, accuracy and detection range on a variety of appliances to demonstrate its effectiveness towards augmenting NILM systems. We show that accurately detecting 100W loads from 10cm away is possible while maintaining multiple-year battery life from a coin-cell battery.

Event detection for Non Intrusive load monitoring

Monitoring electricity consumption in the home is an important way to help reduce energy usage and Non-Intrusive Load Monitoring (NILM) techniques are a promising approach to obtain estimates of the electrical power consumption of individual appliances from aggregate measurements of voltage and/or current in the distribution system. In this paper, we discuss event detection algorithms used in the NILM literature and propose new metrics for evaluating them. In particular, we introduce metrics that incorporate information contained in the power signal instead of strict detection rates. We show that this information is important for NILM applications with the goal of improving appliance energy disaggregation. Our work was carried out on a publicly-available week-long dataset of real residential power usage.

User-Centered Nonintrusive Electricity Load Monitoring for Residential Buildings

This paper presents a nonintrusive electricity load-monitoring approach that provides feedback on the energy consumption and operational schedule of electrical appliances in a residential building. This approach utilizes simple algorithms for detecting and classifying electrical events on the basis of voltage and current measurements obtained at the main circuit panel of the home. To address the necessary training and calibration, this approach is designed around the end-user and relies on user input to continuously improve its performance. The algorithms and the user interaction processes are described in detail. Three data sets were collected with a prototype system (from a power strip in a laboratory, a house, and an apartment unit) to test the performance of the algorithms. The event detector achieved true positive and false positive rates of 94 and 0.26%, respectively. When combined with the classification task, the overall accuracy (correctly detected and classified events) was 82%. The advantages a...

Training load monitoring algorithms on highly sub-metered home electricity consumption data

Abstract The growing interest in energy-efficient buildings is driving changes in investment, design, and occupant behavior. To better focus cost and resource conservation efforts, electricity consumption feedback can be used to provide motivation, guidance, and verification. Disaggregating by end-use helps both consumers and producers to identify targets for conservation. While hardware-based sub-metering is costly and labor-intensive, non-intrusive load monitoring (NILM) is capable of gathering detailed energy-use data with minimal equipment cost and installation time. However, variations in measurements between metering devices complicate the process of compiling the necessary appliance profiles. Future work involves the development of NILM algorithms using sensor fusion and detailed appliance-level data gathered from a highly-sensed house currently being constructed near Pittsburgh, Pennsylvania.

Sensing and Field Data Capture for Construction and Facility Operations

Collection of accurate, complete, and reliable field data is not only essential for active management of construction projects involving various tasks, such as material tracking, progress monitoring, and quality assurance, but also for facility and infrastructure management during the service lives of facilities and infrastructure systems. Limitations of current manual data collection approaches in terms of speed, completeness, and accuracy render these approaches ineffective for decision support in highly dynamic environments, such as construction and facility operations. Hence, a need exists to leverage the advancements in automated field data capture technologies to support decisions during construction and facility operations. These technologies can be used not only for acquiring data about the various operations being carried out at construction and facility sites but also for gathering information about the context surrounding these operations and monitoring the workflow of activities during these o...

Robust ultrasonic damage detection under complex environmental conditions using singular value decomposition.

Guided wave ultrasonics is an attractive monitoring technique for damage diagnosis in large-scale plate and pipe structures. Damage can be detected by comparing incoming records with baseline records collected on intact structure. However, during long-term monitoring, environmental and operational conditions often vary significantly and produce large changes in the ultrasonic signals, thereby challenging the baseline comparison based damage detection. Researchers developed temperature compensation methods to eliminate the effects of temperature variation, but they have limitations in practical implementations. In this paper, we develop a robust damage detection method based on singular value decomposition (SVD). We show that the orthogonality of singular vectors ensures that the effect of damage and that of environmental and operational variations are separated into different singular vectors. We report on our field ultrasonic monitoring of a 273.05 mm outer diameter pipe segment, which belongs to a hot water piping system in continuous operation. We demonstrate the efficacy of our method on experimental pitch-catch records collected during seven months. We show that our method accurately detects the presence of a mass scatterer, and is robust to the environmental and operational variations exhibited in the practical system.

Data-Fusion Approaches and Applications for Construction Engineering

Data fusion can be defined as the process of combining data or information for estimating the state of an entity. Data fusion is a multidisciplinary field that has several benefits, such as enhancing the confidence, improving reliability, and reducing ambiguity of measurements for estimating the state of entities in engineering systems. It can also enhance completeness of fused data that may be required for estimating the state of engineering systems. Data fusion has been applied to different fields, such as robotics, automation, and intelligent systems. This paper reviews some examples of recent applications of data fusion in civil engineering and presents some of the potential benefits of using data fusion in civil engineering.

Using smart devices for system-level management and control in the smart grid: A reinforcement learning framework

This paper presents a stochastic modeling framework to employ adaptive control strategies in order to provide short term ancillary services to the power grid by using a population of heterogenous thermostatically controlled loads. The problem is cast anew as a classical Markov Decision Process (MDP) to leverage existing tools in the field of reinforcement learning. Initial considerations and possible reductions in the action and state spaces are described. A Q-learning approach is implemented in simulation to demonstrate how the performance of the new MDP representation is comparable to that of a Linear Time-Invariant (LTI) one on a reference tracking scenario.