Z-World	Desktop	Wolf BL2600
A Controller Based Networked Autonomous Weather Server
S. Cambeşteanu, M.A. Panait, and A.M. Morega
Faculty of Electrical Engineering, POLITEHNICA University of Bucharest, Romania
[email protected], [email protected], [email protected]
Abstract
The paper describes a SBC-based weather server.
Its main advantages over traditional architectures
are greater flexibility, lower costs, and higher
reliability in the field. By using a proper enclosure,
the weather server can be operated outdoors in
remote locations, and it can provide information on
the main weather parameters such as wind direction
average and maximum speed, atmospheric pressure
momentary and maximal for the day, temperature,
illumination, short term forecasts, etc., in the form of
an easy to read webpage that also allows for
commands to be passed to the weather server. A
networked weather server operates in a distributed
monitoring system, covering large geographical
areas and helping forecasts with local data.
1. Introduction
A weather server consists of several transducers
and gauges, aimed at monitoring weather data,
organized as an autonomous weather station around
a network-enabled controller, capable of reading,
interpreting and delivering the results in form of
webpage or specially formatted e-mails. This paper
discusses the possibility of building a simple and
reliable weather server based on the Z-World "Wolf"
single board computer [1].
The system we built is based on the Single Board
Computer (SBC) concept. More than a controller, a
SBC has also various input and output modules and a
network card (LAN adapter), supporting the
following interfaces: Ethernet, PPP, and PPP over
Ethernet. The programming language for the
controller is Dynamic C, a feature-rich, PC-based
computer programming language that allows for
complex programming to be carried out on a PC and
then loaded into the SBC's EEPROM via a
programming cable. The paper discusses the
possibility of building a simple and reliable weather
server based on the Z-World "Wolf" SBC [1].
2. Project Description
A weather monitoring instruments comprises
atmospheric pressure gauges, temperature, light,
humidity, and atmospheric turbidity sensors.
Commonly, the software that interprets the signals
from these sensors will then provide for short term
weather forecasts based on corroborative usage of
data, and will make accessible these analyses upon
request, along with the regular instrument readouts.
The weather station we developed is equipped
with a thermometric sensor provided with a
thermistor and a fixed resistor, a piezoelectric
pressure gauge, an anemometer for wind speed and
direction measurements, and a diffuse luminosity
sensor. These sensors have output signals in the
0...+5 V range, and can be directly connected to the
inputs of the Z-World SBC. The temperature sensor
uses the thermistor to build a voltage divider; the
temperature is read across the thermistor and
multiplied by a tabulated constant (here, the
linearization of the time constant is performed by the
table method).
This solution has the advantage of simplicity and
reduced cost, as compared to a dedicated sensor.
However, the linear LM35 dedicated sensor would
be better option if higher precision is required. The
resistive potential divider setup has the advantage of
a greater thermal constant thus rendering it relatively
immune to rapid variations due to special conditions,
by reading a mean value of the temperature instead.
The pressure sensor is a dedicated piezoelectric type
absolute pressure sensor. The strongly non-linear
output signal, typical for these transducers, is
linearised by the program by using a constant table
from the product's documentation.
The wind speed sensor uses two-disk slot optical
encoders. One of them is for wind speed, having a
disk with a single slot and fixed optical detection
sensor. The second one consists of two discs, one
with two rows of slots, for marking the true North
and the other one for counting angular position
increments, and the other having a single fixed slot,
oriented by the true North. An "and" logic gate
provides for the reset signal to a reversible counter
0
when the total angle exceeds 360 and will also
change the counting direction.
As a starting point in this project we use a simple
solar tracker consisting of a two axis mobile mount
€
powered by stepper motors, which actuates a small
solar panel, and possibly a few instruments such as a
diffuse radiation sensor. The tracker is included in
the autonomous weather station as support for the
luminosity
sensor
and
direct
insolation
measurements – it is intended for usage in a small
solar farm.
The tracker is equipped with a special neural
network that enables its autonomous operation.
However, in normal operation, it is periodically
bypassed, and control is entrusted to the SBC
controller for raw positioning, then it is restored to
the neural network to make finer position
adjustments.
Next we analyze the structure of this application:
The system is designed around the Z-World BL2600
Wolf controller. A web based application handles the
main parameters display like: air temperature,
uptime wind speed, wind direction, relative
humidity, luminosity, energy converted by the panel
etc. The main functions are presented next.
Panel positioning control. The panels are
positioned by using the coordinates computed in the
program, using the Equation of Time. This equation
has the current date and time as input, and gives the
Sun's position in the sky at a certain moment, and
two angular coordinates as output. The program
transforms these coordinates into steping sequences,
for each stepper motor on each axis.
Data logging. Data provided by the sensors is
acquired periodically and stored in the flash memory,
within the controller. Later, it may be downloaded
through the controller’s web page. Also the data log
can be sent by the end of each day, through e-mail.
Error handling. The controller is capable to
analyze the difference between the calculated
coordinates and the correction made by the neural
network. If this difference is large enough (a
threshold of 1 to 4 degrees is specified, depending on
the motors and gears), the controller overwrites the
correction, reports the error via e-mail, and it follows
the computed trajectory that matches the sun position
at the specified date. Besides, it monitors all other
parameters such as the panel temperature, which
must be lower than the maximum value specified by
the panel spec.
The core component of the system is a BL2600
controller. The BL2600 is a high-performance, Cprogrammable single-board computer that offers
built-in digital and analog I/O combined with
Ethernet connectivity in a compact form factor. The
BL2600 is ideal for both discrete manufacturing and
process-control applications. It is an advanced
single-board computer that incorporates a powerful
Rabbit 3000 microprocessor operating at 44.2 MHz;
512K static RAM and 512K flash memory standard;
36 digital I/O: 16 protected digital inputs, 4 highcurrent digital outputs software configurable as
sinking or sourcing, and 16 I/O individually
software-configurable as inputs or sinking outputs;
12 analog channels: eight 11-bit A/D converter
inputs, four 12-bit D/A converter 0-10; V or ±10 V
buffered outputs; one RJ-45 Ethernet port compliant
with IEEE 802.3 standard for 10/100Base-T Ethernet
protocol; three Ethernet status LEDs; three serial
ports (2 RS-232 or 1 RS-232 with RTS/CTS, 1 RS485 or RS-232); two RabbitNet™ expansion ports;
battery-backed real-time clock; watchdog supervisor.
The sensors comprise either a resistive
temperature sensor or a dedicated LM35 chip with a
greater sensitivity and a linear output (Fig. 1), a
piezoelectric pressure gauge, a small solar panel, a
diffuse luminosity sensor consisting in several
photodiodes with suitable filters, lens and opaque
shades, an anemometer with wind direction
indicator, and a relative humidity capacitive sensor
(sensitive and having a linear output).
Figure 1. The temperature sensor, based on a 10 Ohm NTC
thermistor and a precision resistor. Note that the resistor is
near the controller input, and the thermistor is farther away
(30 cm of cable) to minimize common mode errors due to
the heating of both components.
The solar tracker (Fig. 2) is made of two stepping
motors, one two-axis frame, a neural controller and
three state buffers for allowing the controller to
assume command, fixtures for the solar panel and
instruments.
Figure 2.Tracker mount. It is a one axis, adjustable mount
for driving a small solar panel (0.25W) used for
measurements and tracking efficiency assessment.
The neural network (Fig. 3, 5) simplifies the task
of the SBC, as the tracker and the processing unit are
entirely controlled by it. Its only input is an “enable”
connection, which starts and stops the motor (Fig. 5)
driving square-wave pattern generation. The neural
network generates the position control and the fuzzy
logic commands under the direct influence of the
light gradient sensors.
Figure 3. The signal generator used in testing the neural
network, built around an EXAR 8038 precision waveform
generator chip and momentarily powered from an external
filtered tracking supply (not shown).
Figure 4. The neural network mounted on a quick-board
(no soldering) undergoes tests for stability assessment.
Figure 5. Stepping motor interface (power buffer).
The software. A dedicated application is
developed for the reading of time and date, the
calculation of the most probable position of the Sun,
for recording various parameters such as panel
temperature, voltage output, current output, etc., and
for either broadcasting over the Internet or
displaying them via a web page.
The system’s structure is depicted in Fig. 6. The
diagram shows the dual-control strategy: the neural
network drives the tracker on two axes (X and Y) for
orientation and position adjustment, respectively.
The SBC pre-positions the tracker according to the
Equation of Time for the current moment so that the
tracking motions are reduced to a minimum. Other
sensors connect to the SBC, including the tracker’s
own encoders that allow the Sun’s position in the sky
to be calculated and a direct illumination sensor can
be read, along with a diffuse illumination one, to
characterize the atmospheric turbidity factor.
Figure 6. System functional diagram.
BL2600 has a built-in Ethernet LAN adapter,
which can be internetworked with other elements.
This is a very important feature because data, such as
the status of the application or the weather
parameters, can be accessed through Internet. The
supported interfaces are: a) Ethernet; b) PPP (Pointto-Point Protocol) over a serial link; c) PPP over
Ethernet.
The available interfaces depend on the hardware
configuration of the target board. If an Ethernet port
is available, then it may be used for normal Ethernet
or PPP over Ethernet (PPPoE) connection. The
TCP/IP protocol is implemented in Dynamic C,
which helps developing C language applications for
Rabbit controllers. It includes several utility libraries.
The main library is DCRTCP.LIB. As of
Dynamic C 7.05, this library is a light wrapper
around DNS.LIB, IP.LIB, NET.LIB, TCP.LIB and
UDP.LIB. These libraries implement DNS (Domain
Name Server), IP, TCP, and UDP (User Datagram
Protocol). This, along with the libraries ARP.LIB,
ICMP.LIB, IGMP.LIB and PPP.LIB are the transport
and network layers of the TCP/IP protocol stack.
Dynamic C 7.30 offers support also for a second
Ethernet, should such a board becomes available.
To run the TCP/IP stack, the host (controller
board) needs to know its unique home IP address for
each interface. Interfaces that connect to broadcast
networks (i.e., Ethernet) must also have a netmask
assigned. The combination of IP address and
netmask describes the so-called subnet, which is
addressable via that interface. The subnet basically
describes the community of host addresses that can
talk directly to this host, without requiring data to
pass through a packet router. Point-to-point links
need an IP address only, as single hosts.
The network configuration information like IP,
MASK, DNS, Gateway, and addresses are stored in
the library called "dcrtcp.lib", either static or
dynamic: static by using predefined information
inside the lib file, and dynamic by enabling the
DHCP function (also inside of the lib file).
The program is very flexible: it recognizes
commands like “ifconfig” for setting up the network
interface parameters (similar to UNIX), and can also
configure the PPP interfaces, usually a slightly more
complicated task than for the non-PPPoE Ethernet.
The advantage of using PPP is that it can run over a
wide variety of physical layer hardware: on Rabbitbased boards it includes the asynchronous serial
ports, the Ethernet (PPPoE), and PPP over
asynchronous serial allows boards with no Ethernet
hardware to communicate using TCP/IP protocols.
PPPoE is often considered a "hack." It seems
superfluous to define a protocol that establishes a
logical "connection" between two peers on what is
otherwise a broadcast (i.e., any-to-any) medium.
Nevertheless, the existence of PPPoE was largely
dictated by the needs of ISPs who wished to continue
using their existing infrastructure, based on the
earlier generation of dial-in connections. The
availability of high-speed (ADSL, etc.) modems,
with Ethernet connection to the user's network,
makes an attractive alternative out of PPPoE. If the
application requires connection to an ISP via an
ADSL modem, then most likely, PPPoE support is
needed.
PPPoE requires a physical layer negotiation to
precede the normal PPP negotiations. This is known
as the "access concentrator discovery" phase
("discovery" for short). PPPoE makes a distinction
between PPPoE servers and PPPoE clients. However
PPP makes no distinction; PPP may be seen as also
standing for Peer-to-Peer Protocol. The PPPoE
server is known as the access concentrator. The
Dynamic C TCP/IP libraries do not support acting as
the access concentrator; only the PPPoE client mode
is supported. This is the most common case, since
the DSL modem is always configured as an access
concentrator.
The controller also supports web applications
such as HTTP server, hosting for sites or FTP server.
Most users of the application will be familiar with at
least one web browser (e.g., Mozilla, Internet
Explorer, Opera, Safari), and they can use the web
site as a graphical interface to operate the controller.
In our case we use the controller to provide weather
information over the Internet. Then, information may
be anonymously accessed. By utilizing the controller
any format and layout of the web pages that will be
presented to the browser can be implemented.
Should the application return information only,
without allowing for updates (such as a data logger)
then forms may be used. Forms, in web parlance,
allow the browser's user to fill in some information
then submit it to the server. The server, on the
application, then performs the requested actions and
sends a confirmation back to the browser. This is the
most common means for implementing control of the
server as opposed to merely querying it.
3. Results
3.1. The WebServer architecture
The server is responsible for fielding requests
from the outside world. Each request is analyzed to
determine the resource that is requested, the user
who makes the request, and whether the user is
authorized to obtain that resource.
Table 1. The software routine for a webpage displaying
data read from the weather server’s instruments.
If the resource is available, the user is known and has
the proper permissions then the resource is made
available via the browser. A software routine for
exporting the data is presented in Table 1.
Figure 7. A LabView interface offers an intuitive picture of
the measured weather variables.
The user interface/input module can be hosted,
for instance, on a LabView server that allows for a
simple yet effective GUI (Fig. 7). The interface
displays the main instrument’s readings, and can
support user commands to query or calibrate the
system.
field data, to aid simulations and forecasts. Modern
weather forecast systems are based on such
distributed weather-monitoring network, including
airport weather stations and small private weather
stations as nodes.
This paper reports a simple and reliable weather
server based on the Z-World "Wolf" single board
computer, equipped with transducers and gauges
capable of monitoring weather data, organized as an
autonomous weather station around a networkenabled controller, capable of reading, interpreting
and delivering results in the form of webpage or
specially formatted e-mails. The server is conceived
as the elemental cell of a distributed internetworked
system, aimed at assisting weather monitoring and
forecast, or as part of distributed energy generation
systems – e.g., photovoltaic, wind, tidal power
farms, etc.
Acknowledgements
Figure 8. Networked array of webservers.
Finally, Fig. 8 shows a networked array of peerto-peer weather servers that cooperate in providing
regional meteo and weather data. A specialized
server (to be implemented at iem.pub.ro) would then
concentrate and process the acquired data, for
storage and further usage.
4. Conclusions
Recent developments in weather monitoring show
the paramount importance of such distributed
weather stations, connected in a network capable of
covering large geographical areas and delivering in-
Panait M. gratefully acknowledges the support
offered by the CNCSIS grant TD /2007-2008 and by
the IEM Laboratory.
References
[1] BL2600 Wolf Reference Manual, Z-World,
http://www.zworld.com (accessed in July, 2008).
[2] I.A. Morega, C. Mogos, M.A. Panait, Introducere in
programare pentru controlul integrat, MatrixRom,
2006.
[3] Wikipedia: Equation of Time
[4] R. Panko, “Business Data Networks and
Telecommunications”, 6th Ed, Prentice-Hall, 2007
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF

advertisement