Schneider Electric (sponsoring 1 award)
Topics of Interest:
Residential Energy Monitoring
Energy monitoring from residential wiring devices and communicating information to a central monitoring system located at the main electrical service location. The objective is to provide detailed information on energy usage and potentially other electrical characteristics such as power quality. Existing functional products will be provided to assist in the research activity. (This could be an expansion of the work Rosemary Alden has done)
Integration of PV into residential electrical service applications as an alternative energy source and including the necessary Energy Storage System. The ability to monitor energy provided for each source as well as provide the homeowner with the ability to manage the transfer themselves or make it automatically.
Harbor System Automation
Research the use of automation/control systems to improve the electrification of Harbors/Port systems to improve energy efficiency, increase in cargo throughput, and improve safety.
North American Stainless (sponsoring up to 2 awards)
Topics of Interest:
Develop a monitor for active factory signals – a system that can run on all 40,000+ tags all of the time. This will be based on a neural network/relationship matrix that will detect some deviations. Active Factory SQL currently has built-in signal health values used to monitor. This will be the detection of abnormal signals - both standalone detection, and relational limits with multiple tags.
Example 1: thermocouple which normally reads 2050F on normal run conditions, reads 2250F constantly – flag for bad Input condition. In this case, AFSQL is getting valid data, but data is incorrect.
Example 2: Compare multiple pyrometer signals on a process line – for process speed feedback limits, downstream pyros should be within a calculated range. Alarm when out of range.
Closed-Loop Process Monitoring
Develop generalized logic to detect issues with closed-loop systems. This will involve looking at basic PID input/output conditions, determining normal boundaries, setting limits, and detecting problems.
This should be a general algorithm to be applied to any control signals available; specifically speeds, tensions, combustion parameters, and water treatment.
Example 1: Zone x of furnace has Temperature SP 2210F, Temp_fdbk 1975F, Gas_valve_pos = 95%. Alarm for IVP max without reaching SP.
Example 2: Load cell feedback shows oscillation during steady-state conditions of process line (speeds, temperatures, etc). Create an alarm for a bad sensor/process parameter.