Warit Wichakool

Smart Metering of Variable Power Loads

Nonintrusive load monitoring (NILM) seeks to determine the operation of individual loads in a building strictly from measurements made on an aggregate current signal serving a collection of loads. Great strides have been made in performing NILM for loads whose operating state can be represented by a finite-state machine, i.e., loads that consume discrete or distinct power levels for periods of time. It is much more difficult to track the operation of continuously variable loads that demand ever-changing power. These loads are becoming more prevalent as variable speed drives, daylight-responsive lighting, and other power electronic controlled loads emerge on the grid. This paper demonstrates a method for tracking the power consumption of variable demand loads nonintrusively. The method applies to any site where NILM might be of interest, including commercial and industrial buildings, residences, and transportation systems.

Fault detection and diagnostics for non-intrusive monitoring using motor harmonics

Harmonic analysis of motor current has been used to track the speed of motors for sensorless control. Algorithms exist that track the speed of a motor given a dedicated stator current measurement, for example [1–5]. Harmonic analysis has also been applied for diagnostic detection of electro-mechanical faults such as damaged bearings and rotor eccentricity [6–17]. This paper demonstrates the utility of harmonic analysis for fault detection and diagnostics in non-intrusive monitoring applications, where multiple loads are tracked by a sensor monitoring only the aggregate utility service. An optimization routine is implemented to maintain accuracy of speed estimation while using shorter lengths of data.

Magnetic endotracheal tube imaging device

This paper describes an accurate, economical, and portable device that helps to locate the position of an endotracheal tube (ETT) in situ. The device uses an array of magnetic field sensors to detect an anomaly in magnetic field caused by magnet embedded near the cuff of an ETT, and displays an intuitive map of relative magnetic field intensity under the sensor area. The device provides real-time feedback of the position to a clinician, so that corrective measures can be taken if the ETT is determined to be outside of normal positioning with respect to the patients airway. Variations of the proposed design are suitable for continuous monitoring.

Resolving Power Consumption of Variable Power Electronic Loads Using Nonintrusive Monitoring

This paper demonstrates a new estimator to resolve the average power consumption of a variable-speed drive (VSD) from other constant power loads using a spectral estimation method and a switching function technique for nonintrusive load monitoring (NILM) applications. The estimator uses the rectifier switching function to create adaptive estimating functions for the fundamental power from a finite set of current harmonics generated by the VSD. Experimental results show that the proposed VSD power estimator can significantly improve NILMs ability to resolve the VSD power consumption under small input voltage variation.

High-resolution physically-windowed sensors for power electronics applications

This paper presents a high-resolution, physically-windowed sensor architecture that is well-suited for energy scorekeeping and diagnostic applications. The sensor can track a large-scale main signal while capturing small-scale variations. The prototype system measures a small current signal using a closed-loop Hall sensor, and extends the range by driving a compensation current through an auxiliary winding. The system combines the compensation command and the sampled output of the residual sensor to reconstruct the input signal. Results show that the prototype can measure both dc and ac currents with 10 mA resolution over a 160 A current range.

A waveform-based power estimator for variable power loads

This paper proposes a method to derive an estimator that predicts the power consumption of variable power loads from a subset of higher current harmonics without requiring a full analysis of the internal circuit of the load. The method exploits structural features of the current waveforms consumed by the load to develop the estimator. The computation involves Gaussian elimination of a cyclotomic field representation to compute the estimator coefficients, avoiding floating-point computational error. Experimental results have shown that the proposed algorithm can derive estimators that can extract the power consumption of variable speed drives, computers, or light dimmers from fixed power loads in aggregate measurements.