Transcript Energy Tracking & Storage for an Autonomous System (ETSAS)

Energy Tracking &
Storage for an
Autonomous System
OASIS
Mir Ziyad Ali
Liron Kopinsky
Christopher Wallace
Sarah Whildin
Joseph Yadgar
Overview
 We
are building an energy tracking
system that will position solar panels
in order to get maximum power.
 Examples
1. Mars Rover
2. Hubble Telescope
3. International Space Station
4. Solar-Powered Homes
Main Objectives




Position the Solar Panels so that they will acquire
maximum energy from a light source.
Store the acquired energy into batteries, and use
the batteries to control the rest of the system
when the solar energy is absent.
In no-light or low-light conditions design the
system to go into sleep mode so that energy is
not wasted.
Design a visual display unit to display status
information about the system.
Project Block Diagram
Possible Extensions
Controlling the system via a wireless
transceiver.
 The possibility to reprogram the FPGA
from another location.
 Equipping the system with a remote
manual override for solar panel positioning
in the event of a malfunction.
 Data acquisition of power readings and
solar power distribution to analyze light
patterns .

Risks and Contingencies
 Insufficient power is collected from the
light source to operate the control system.
 Implement low-light intensity control system
(power-save mode).
 Cost of high-efficiency solar panels is too

high.
Use smaller sized solar panels.
Project Schedule
Division of Labor
Mir Ziyad Ali: Battery Charging and Power
Management
 Liron Kopinsky: FPGA hardware
design/software
 Chris Wallace: FPGA hardware
design/software
 Sarah Whildin: Solar panel movement and
control
 Joseph Yadgar: Digital to Analog
Component Integration

Power Management
Objectives:
 Design a High Efficient Power
Source for the entire ETSAS
digital System.
 Solar Energy will be converted to
DC Voltage, which will be
regulated and supplied for the
Microcontroller.
SOLAR PANELS
CHARGE CONTROLLER
DEEP CYCLE BATT
MICROCONTROLLER
Solar Panels




Siemens ST Series multiplelayer CIS solar cells with 12V
Output.
Solar panels will be directed
at solar south in the northern
hemisphere.
Boulder, CO has 5.72 solar
insulation in kilowatt-hours
per square meter per day.
Three Solar panels will be
connected in series to
produce 60 Watts (17.1V x
3.5 A) of total power.
Charge Controller/Regulator


A charge controller monitors the
battery's state-of-charge to
insure that when the battery
needs charge-current it gets it,
and also insures the battery is not
over-charged.
This controller will use the
charging principal called the
Maximum Power Point
Tracking (MPPT) which
maximizes the amount of current
going into the battery from the
solar array by lowering the
panel's output voltage.
Deep Cycle Battery


The Deep Cycle batteries will
be used as they are designed
to be discharged and then recharged for thousands of
times.
Sealed no-maintenance
versions Lead-Acid Ah
batteries are the most
common in PV (Photo Voltaic)
systems because their initial
cost is lower.
Movement Design
 We
will use two motors:
- 360 degree horizontal motion
- approximately 270 degree vertical
motion
Motors
Hybrid unipolar
stepper motors will
allow for high
torque and step
rates.
 Steps will be 1.8°
apart so that 200
steps makes a
complete rotation.


Stepper Motors - use
an electromagnet to
rotate the rotor a
certain number of
degrees.
Motor Control System
 Converts
signals from the
microprocessor detailing direction
and step length and converts them
to motion on the motor.
 Purchase vs. Build
Hardware

Digilent Spartan-3 Board
– 1MB onboad SRAM
– 2Mb Platform Flash
– Switches and Pushbuttons for Manual Control of Solar
Panels and LCD contents.
– Xilinx Spartan3 XC3S200-FT256


Small LCD for status information
Miscellaneous
– Level Shifters
– A/D converters for Solar Panel energy readings and
battery level
– 4 layer PCB
– Other stuff we haven’t found out we need.
Inside the FPGA







Microblaze 32-bit Microprocessor (~50% FPGA)
50MHz Max on Digilent Board (~40 DMips)
27K Internal BRAM for used with OPB BRAM
controller core
JTAG_UART IP core + Xilinx Microprocessor
Debugger + GDB debugger (allows software
debugging capabilities)
Undetermined amount of GPIO ports for
interfacing with the rest of the system.
OPB Interrupt Controller for use with timers
Any custom logic such as decode logic to
communicate with LCD if it is memory-mapped,
etc.
Tracking algorithm
 Considered
using a basic
search algorithm
 Instead decided to use
multiple solar panels and to
take differentiation readings
 This will allow for us to
detect in which direction/s
the light intensity is
stronger
Tracking continued
 Unfortunately,
our search algorithm
fails if the only light is directly behind
the panels
 To fix this problem, we intend to put
three small stationary panels around
the base of the module
 We will then use a triangulation
algorithm to determine in which
direction the light source is
Logic implementation
 Using
the Xilinx Embedded
Development Kit (EDK) we intend to
program most of our logic using
C/C++
– Additional code optimizations necessary
will then be programmed in assembly if
necessary
 Additional
logic necessary on the
FPGA (e.g. for data retrieval) will be
programmed in Verilog
Advantages Over Existing Systems
By providing a small, modular design, we
provide for many different panel
configurations not allowed on larger-scale
designs
 By remaining small, our system should be
much more power efficient than larger
moving solar arrays
 By allowing for system reprogrammability, we ensure that our
system will be able to be optimized further
with, for example, new weather data
