Engineering topics

Castello, Charles C

Publications and source records attributed to Castello, Charles C.

6 records

DOE OSTI2015

Gamification and Visualization of Sensor Data Analysis in Research Buildings

The use of video game elements in non-gaming systems, or gamification , has potential value in transforming data analysis. Our study focused on creating a web-based videogame that models two physical test buildings, each of which contains hundreds of sensors. After the application renders the models, the player can walk through the environments and interact with the virtual representations of the sensors inside. Rather than trudging through a database with textual commands and screens full of data, the user can (virtually) walk up to a sensor and view its data graphically. But these features only scratch the surface of what is possible using our new gamification approach. We anticipate being able to show that recent progress in game design techniques and capacities can contribute to the field of analysis through gamification. The net result could be more stimulating, intuitive, user-friendly interfaces, as well as potentially more informative and insightful applications.

Stone, Jackson A↗

DOE OSTI2014

Sensor Data Management, Validation, Correction, and Provenance for Building Technologies

Oak Ridge National Laboratory (ORNL) conducts research on technologies that use a wide range of sensors to develop and characterize building energy performance. The management of high-resolution sensor data, analysis, and tracing lineage of such activities is challenging. Missing or corrupt data due to sensor failure, fouling, drifting, calibration error, or data logger failure is another issue. This paper focuses on sensor data management, validation, correction, and provenance to combat these issues, ensuring complete and accurate sensor datasets for building technologies applications and research. The design and development of two integrated software products are discussed: Sensor Data Validation and Correction (SensorDVC) and the Provenance Data Management System (ProvDMS) platform.

Data management↗

DOE OSTI2014

Control Strategies for Electric Vehicle (EV) Charging Using Renewables and Local Storage

The increase of electric vehicle (EV) and plug-in hybrid-electric vehicle (PHEV) adoption creates a need for more EV supply equipment (EVSE) infrastructure (i.e., EV chargers). The impact of EVSE installations could be significant due to limitations in the electric grid and potential demand charges for residential and commercial customers. The use of renewables (e.g., solar) and local storage (e.g., battery bank) can mitigate loads caused by EVSE on the electric grid. This would eliminate costly upgrades needed by utilities and decrease demand charges for consumers. This paper aims to explore control systems that mitigate the impact of EVSE on the electric grid using solar energy and battery banks. Three control systems are investigated and compared in this study. The first control system discharges the battery bank at a constant rate during specific times of the day based on historical data. The second discharges the battery bank based on the number of EVs charging (linear) and the amount of solar energy being generated. The third discharges the battery bank based on a sigmoid function (non-linear) in response to the number of EVs charging, and also takes into consideration the amount of renewables being generated. The first and second control systems recharge the battery bank at night when demand charges are lowest. The third recharges the battery bank at night and during times of the day when there is an excess of solar. Experiments are conducted using data from a private site that has 25 solar-assisted charging stations at Oak Ridge National Laboratory (ORNL) in Oak Ridge, TN and 4 at a public site in Nashville, TN. Results indicate the third control system having better performance, negating up to 71% of EVSE load, compared with the second control system (up to 61%) and the first control system (up to 58%).

Castello, Charles C↗

DOE OSTI2013

Minimization of Impact from Electric Vehicle Supply Equipment to the Electric Grid Using a Dynamically Controlled Battery Bank for Peak Load Shaving

This research presents a comparison of two control systems for peak load shaving using local solar power generation (i.e., photovoltaic array) and local energy storage (i.e., battery bank). The purpose is to minimize load demand of electric vehicle supply equipment (EVSE) on the electric grid. A static and dynamic control system is compared to decrease demand from EVSE. Static control of the battery bank is based on charging and discharging to the electric grid at fixed times. Dynamic control, with 15-minute resolution, forecasts EVSE load based on data analysis of collected data. In the proposed dynamic control system, the sigmoid function is used to shave peak loads while limiting scenarios that can quickly drain the battery bank. These control systems are applied to Oak Ridge National Laboratory s (ORNL) solar-assisted electric vehicle (EV) charging stations. This installation is composed of three independently grid-tied sub-systems: (1) 25 EVSE; (2) 47 kW photovoltaic (PV) array; and (3) 60 kWh battery bank. The dynamic control system achieved the greatest peak load shaving, up to 34% on a cloudy day and 38% on a sunny day. The static control system was not ideal; peak load shaving was 14.6% on a cloudy day and 12.7% on a sunny day. Simulations based on ORNL data shows solar-assisted EV charging stations combined with the proposed dynamic battery control system can negate up to 89% of EVSE load demand on sunny days.

Battery Management Systems↗

DOE OSTI2013

Autonomous Correction of Sensor Data Applied to Building Technologies Using Filtering Methods

Sensor data validity is extremely important in a number of applications, particularly building technologies where collected data are used to determine performance. An example of this is Oak Ridge National Laboratory s ZEBRAlliance research project, which consists of four single-family homes located in Oak Ridge, TN. The homes are outfitted with a total of 1,218 sensors to determine the performance of a variety of different technologies integrated within each home. Issues arise with such a large amount of sensors, such as missing or corrupt data. This paper aims to eliminate these problems using: (1) Kalman filtering and (2) linear prediction filtering techniques. Five types of data are the focus of this paper: (1) temperature; (2) humidity; (3) energy consumption; (4) pressure; and (5) airflow. Simulations show the Kalman filtering method performed best in predicting temperature, humidity, pressure, and airflow data, while the linear prediction filtering method performed best with energy consumption data.

Castello, Charles C↗

DOE OSTI2013

Machine Learning Techniques Applied to Sensor Data Correction in Building Technologies

Since commercial and residential buildings account for nearly half of the United States' energy consumption, making them more energy-efficient is a vital part of the nation's overall energy strategy. Sensors play an important role in this research by collecting data needed to analyze performance of components, systems, and whole-buildings. Given this reliance on sensors, ensuring that sensor data are valid is a crucial problem. Solutions being researched are machine learning techniques, namely: artificial neural networks and Bayesian Networks. Types of data investigated in this study are: (1) temperature; (2) humidity; (3) refrigerator energy consumption; (4) heat pump liquid pressure; and (5) water flow. These data are taken from Oak Ridge National Laboratory's (ORNL) ZEBRAlliance research project which is composed of four single-family homes in Oak Ridge, TN. Results show that for the temperature, humidity, pressure, and flow sensors, data can mostly be predicted with root-mean-square error (RMSE) of less than 10% of the respective sensor's mean value. Results for the energy sensor are not as good; RMSE are centered about 100% of the mean value and are often well above 200%. Bayesian networks have RSME of less than 5% of the respective sensor's mean value, but took substantially longer to train.

building technologies↗