CMU Voyager Instrumenting for Pollution and Energy Consumption

CMU–Voyager Instrumenting for Pollution and Energy Consumption

Team Members Pratik Agarwal Alex Eiser Gary Feigenbaum Yuan-Ning “Richard” Hsieh Gregor Kronenberger Kietae Park Asad Samar Kristen Stubbs

Overview

Overview: Goals • Instrument Voyager to monitor —Propulsion power generation —Energy Consumption —Energy Efficiency —Fuel Consumption

Overview: Goals • Provide a means to view this data —For instructors: • Teaching tools —For the crew: • Visualize the information gathered • Provide ability to perform basic analysis • Provide insights for the new Voyager

Overview: System Architecture • Three subsystems —Network of sensors • • • Fuel flow rate Electricity sensors GPS sensor —Scientific visualization tool for the crew —“Dashboard” interface for students

Overview: Design Process • PHASE I: • Understand the problem • Create a visionary scenario • Propose an initial solution PHASE II: • Create a concrete design • Start on implementation • Discover and solve problems • Update design PHASE III: • Complete implementation • System integration and testing • Demonstration

Final Prototype Application Launcher

System Architecture: Application Launcher

System Architecture: Application Launcher Wait until Status turns green Launch Dashboard

Final Prototype Dashboard

Final Prototype Scientific Visualization

Data Visualization: Single Dimension • Web interface to view data • Works in IE 6/ Opera 7. 5/ Mozilla • Requires Javascript for date selection • Shows one-dimensional data vs. time

Data Visualization • Choose Sensor —Speed —Electrical —Fuel consumption —Not pollution

Data Visualization • Select date —using a dropdown —using the calendar

Data Visualization • Select date —using a dropdown —using the calendar

Bar Charts: Fuel Consumption vs. Time

Line Charts: Fuel Consumption vs. Time

Two Sensors: Fuel Consumption and RPM vs. Time

Excel Export Link in graph display • Exports directly to an Excel file for further analysis •

Excel Export Link in graph display • Exports directly to an Excel file for further analysis •

Excel Export Link in graph display • Exports directly to an Excel file for further analysis •

System Architecture Overview

System Architecture

System Architecture Sensors

System Architecture: Device Placement

Electrical

System Architecture: Sensors

Current Sensors • Self-powered so can be left permanently running • Output between 0 -5 V depending on the current flowing through the wire • No need to expose the conductor

Electrical Subsystems Shore Power Generator Main circuit panel (hallway) AC/Lights/Heating Switchover Main distribution Pumps panel MSD Ventilation

Sensor Placement Shore Power Generator Main circuit panel (hallway) A/C Heating Ventilation Flushing Pumps Sink Switchover Main distribution Pumps panel MSD Ventilation

Current Sensors • Experiments, first in lab and then on the boat show they are pretty accurate • All data collected in real time now • Seven electrical sensors (worth approx. $700)

Current Sensor Installation • Current sensors have been installed on the boat, to monitor the following: • A/C: 3 units in the main classroom • Heating: 2 units in the main classroom and 1 in pilot house

Current Sensor Installation • Toilet pump: monitors the toilet • Engine pump: The bilge pump in the control room • Ventilation: Ventilation switch in the control room

Flow Rate Sensor

System Architecture: Flow Rate Sensor

Flow Rate Sensor: Flo. Scan Constantly monitor one engine • Either engine combination can be monitored • Real time data • Will not block fuel line • Can be installed permanently • Powered by 12 V DC • Easily accessible from sensor position • 0 -5 V output • Use Data Acquisition Device to collect data • Approx cost $1, 100 •

Flo. Scan

How Flo. Scan Works Fuel Tank Engine

Flo. Scan Installation • Forward and Return sensors were placed in fuel line • Wires were run to the starboard side under the floor • Wires are not visible for 90 % of path and do not interfere with engines

Flo. Scan Installation • The monitor is mounted on the starboard side wall • Wires run to the 12 V DC power supply and to the DAD

Data Acquisition

Data Acquisition: Data Acquisition Device

Data Acquisition: Data Acquisition Device • 0 -5 V input • RS 232 output • Requires power supply, some interface cards and cables — 11 channels

Data Acquisition: Data Acquisition Device • Experiments in the lab and on the boat have shown that this works fine with the current sensors • 1 device costs $90 with the power supply.

Data Acquisition: Installation • • DAD has been installed on the boat. Input from flow rate and current sensors, output to PC

Data Acquisition: I/O module

Data Acquisition: I/O Module • Specialized Application —Interfaces with DAD & GPS —Communicates with sensors using virtual COM ports through the DAD —Gathers data from sensors and GPS —Stores data in our database —Written in Java

GPS

GPS

GPS: Overview • GPS — Used to track movement of Voyager • • • Distance traveled Speed Location – Longitude, Latitude Able to take accurate measurements Report data in real time USB port used as Virtual Com Port

GPS: Product • Rayming TN-200 — Track up to 12 satellites — Built-in antenna and waterproof — Magnet included for mounting — USB output, NMEA data format

GPS: Implementation • We plan to place the GPS sensor in the open where it can easily communicate with satellites • Testing shows accurate results for speed and location • Data is updated in our database every 10 seconds

Mathematical Model

Mathematical Model

Mathematical Model: Overview RPM as a function of fuel rate - Pollution -

Mathematical Model: RPM vs. Fuel Rate Used one of two engines to get the mathematical model • Turned on the engine at a certain RPM to get the fuel rate and plotted RPM as a function of the fuel rate •

Mathematical Model: RPM vs. Fuel Rate RPM 0. 2 600 0. 5 800 0. 9 900 1. 2 1100 1. 3 1200 1. 4 1300

Mathematical Model: RPM vs. Fuel Rate • From the table above, we derived the following formula • Y = -1812. 5*(X^4) + 6744. 9*(X^3) - 8322. 2*(X^2) + 4325. 9*X + 4. 1796 R 2 = 0. 9993 Where Y = RPM; X = Fuel Rate

Mathematical Model: Pollution Emission Model - In order to understand the emission model, we have to look at the relationship between the RPM and the engine power. - We can derive the engine power value with respect to the value of the RPM when the engine is running •

Mathematical Model: Pollution • RPM vs. Engine Power

Mathematical Model: Pollution Emission model —CO = 1110 * engine power / 174 —HC = 83 * engine power / 174 —NOX = 2190 * engine power / 174 —SO 2 = 310 * engine power / 174 —PM = 69. 948 * engine power / 174 • Output in grams/hour •

System Architecture Software

System Architecture

System Architecture • Sensors • • IO Module • • Application to input data into the database Database • • Data Input from sensor devices Stores both long term and real time data Dashboard Application • Display the real time data

System Architecture: Data Flow Queried every second

System Architecture: Data Flow – Explained • User Input chooses: —Recorded data • • • Toilet flushes Revving the engine Turning on the AC —Real-time data from sensors • Devices connected to the electrical system and engines

System Architecture: Data Flow – Explained • Realtime Data Store —Holds a recent history of sensor activity • Permanent Data Store —Holds the long term history • Interesting Data Store —Holds pre-recorded data • Used to help explain how activity impacts the environment

System Architecture: Dashboard Design

System Architecture: Dashboard Design – Explained • IFLogic —Controls complete application function —Uses a timer to control updating GUI and accessing the database —As data changes in the database, individual components are notified

System Architecture: Dashboard Design – Explained • Observers —Are notified when a database change has occurred —Fetch new data —Perform required conversions —Notify Viewers that new data is available

System Architecture: Dashboard Design – Explained • Viewers —Responsible for displaying Dashboard information —Interface with Observers for fetching data —Perform any analysis

Costs

Costs • Computer (from previous year) —Database —Data Visualization —Dashboard • Electrical sensors: 7 ammeters • Engine sensor: 1 Fuel flow meter • GPS sensor: 1 sensor • Data Acquisition Devices: 1 DAD

Costs Product Task Quantity Cost / Unit Floscan 7000 Fuel flow 1 $944 Floscan monitor Floscan output 1 $188 Veris H 922 Ammeter 7(@ $96) $672 DAD + cable, power supply 0 -5 V measuring device 1 $91 Boat speed 1 $105 Dashboard graphing 1 $40 - - $2040 Rayming GPS sensor JFree. Chart sample code Total

Timeline

Timeline Final Presentation and Demonstration: Today Final Report and User Manual: May 6

Demonstration! On board Voyager