Современные информационные технологии/Компьютерная инженерия

Myasischev AA

Khmelnitsky National University, Ukraine

The use of single-board mini computers and WebIOPi framework for remote access to sensors

For decision-making, for example in tasks control devices we need to have information that can be obtained from sensors. The sensors can be located at a great distance from each other and from the control center. Therefore, to obtain access to sensors we can use Internet.

For remote access to the sensors can be used microcontrollers or mini computers single-board. In this paper, to solve the problem of remote access to the sensors we will use mini computers on single-board. Currently, the most popular are mini computers Raspberry Pi 2, Banana Pi 2 and Orange Pi PC.

Fig.1. Single Board Computers

These computers are united:

1. Small size as a credit card;

2. Quad processor that runs on all computers with frequency 1-1.2 GHz ;

3. RAM 1 GB;

4. SD card instead of the disk to load the operating system and programs;

5. Ethernet port for network connectivity;

6. HDMI output for connecting a monitor or digital TV;

7. USB ports for connection, such as a keyboard, mouse, flash memory;

8. Linux operating system;

9. And most importantly - 40-pin GPIO port that connects devices and sensors, which must be managed.

The main task - is the choice of computer for remote management.

1. Cost (on 02.25.2016, the site http://ru.aliexpress.com with delivery):

- Raspberry Pi 2 - $36.99;

- Banana Pi 2 (BPI-M2 A31S) - $50.21

- Orange Pi pc - $18.99

2. Performance CPU + Memory:

For computational tests with 4 cores:

- Banana Pi 2 (BPI-M2 A31S);

- Orange Pi pc;

- Raspberry Pi 2.

When using one core for computer work (the task is not parallelized):

- Orange Pi pc;

- Banana Pi 2 (BPI-M2 A31S);

- Raspberry Pi 2.

It is noted that Orange Pi runs 3 core, 4-th does not always start.

3. Technical support and the presence of a well-functioning software:

Raspberry Pi 2 - (1); Banana Pi 2 - (2); Orange Pi pc - (3).

Orange Pi has not software for support pins of GPIO port.

For remote control sensors can be used microcontrollers:

1. Arduino Mega256 with Ethernet Shied w5100 - cost $12-15;

2. Arduino nano and network controller enc28j60 - cost $8-9;

3. ESP8266-12 - $2-3;

Experience shows that the microcontrollers well are worked in a local network, but in the global network many packets are lost, and management become unreliable. Mini computers are running the Linux operating system, which has got qualitative network protocols. Therefore mini computer manages sensors and devices better than microcontroller across the WAN. For mini computers can do a high degree of protection to access the managed system. Microcontroller has not resources for nice protocols and protect against hacking. Based on the above, for the remote operation of sensors we will use minicomputer Raspberry Pi 2. As an example, is discussed to connect the pressure gauge and temperature BMP180 to I2C bus of computer. We have to solve the tasks:

- When we are connecting to a computer using a web browser on the screen we must to see the pressure and temperature. Their values must is being changed every 5 seconds;

- When we click on a link of temperature and pressure of the browser should render the graphics temperature and pressure;

- The script on Python must do records of pressure and temperature in the files every 5 minutes. They are used for plotting;

- It is necessary to provide management of device in the absence of his real of IP - address (or DNS name). You only need to connect to the Internet, such as through a standard ADSL modem with installation of the NAT.

We consider the sequence of solving the problem:

1. Installation Raspbian operating system.

We need to copy the image RASPBIAN operating system with the site https://www.raspberrypi.org/downloads/raspbian/, such as a computer running Windows 8.1. And unzip this the file. We need to copy Disk Utility Win32DiskImager with site http://sourceforge.net/projects/win32diskimager and is unzip it. After this we must set the SD card to computer and install on SD card operating system image with help of Disk Utility. After that we set the SD card in the Raspberry Pi computer. We must connect a monitor, keyboard, mouse and an Ethernet cable to the computer. After connecting the power, the computer automatically loads of RASPBIAN and displays the preset menu which is formed of file raspi-config. Options of this file is in the link

https://www.raspberrypi.org/documentation/configuration/raspi-config.md .

2. We must give the computer of Raspberry Pi static IP - address.

Raspberry Pi performs the function of web - server, so it should have a static ip address. For this:

- is changing the contents of the file /etc/network/interfaces to

auto lo

iface lo inet loopback

auto eth0

iface eth0 inet static

address 172.20.0.138

netmask 255.255.0.0

gateway 172.20.200.1

dns-nameservers 8.8.8.8

dhcpcd5 completely is removed from the system by running the command:

sudo apt-get purge dhcpcd5

3. The next step this is to set the framework WebIOPi.

WebIOPi Framework is a software package specifically designed for the Raspberry Pi for remote device management. Together with Raspberry Pi 2, he implements the Internet of Things technology. WebIOPi package allows you to create a variety of custom applications. WebIOPi has the following features:

- Built Web - server implemented in Python;

- Built-In support for more than 30 devices with interfaces UART, SPI, I2C, 1-Wire;

- Javascript /HTML Library which uses for create of a Web-based interface;

- Python/Java libraries which uses for creating applications for Android;

- Supports SoAP protocol for control and interaction between ordinary electronic devices over the network.

WebIOPi has open source, which can be changed by the user. This allows you to increase the number of problems to solve. To customize a package for a specific task we must change the configuration file. In this file we write the pins of GPIO to which are connected of devices. If the sensors are used, they also are written in the configuration file. However it is necessary in some cases to include a device driver (for example BMP180 sensor). We will to install version 0.71 WebIOPi. This new version is supported Raspberry Pi 2, which has 40 pins of GPIO port. For install WebIOPi, we must go the computer through 22 port with help of program putty.exe (login - pi, password - raspberry) and in the terminal to enter the following commands one by one:

$ wget http://sourceforge.net/projects/webiopi/files/WebIOPi-0.7.1.tar.gz

$ tar xvzf WebIOPi-0.7.1.tar.gz

$ cd WebIOPi-0.7.1

Install the patch to work with 40 GPIO Raspberry Pi 2:

$wget https://raw.githubusercontent.com/doublebind/raspi/master/webiopi-pi2bplus.patch

$ patch -p1 -i webiopi-pi2bplus.patch

$ sudo ./setup.sh

To automatically start after reboot WebIOPi we have to execute a command(valid for the image 2015-05-05-raspbian-wheezy.img):

sudo update-rc.d webiopi defaults

For later versions of startup programs is performed as follows:

$ cd /etc/systemd/system/

$ sudo wget https://raw.githubusercontent.com/doublebind/raspi/master/webiopi.service

$ sudo systemctl start webiopi

$ sudo systemctl enable webiopi

Then we have to restart the Raspberry Pi 2: sudo reboot

Now we need to test the WebIOPi. From any computer on the local network, we introduce the network address which has the Raspberry Pi 2 with the port 8000. For example: http://172.20.0.138:8000/app/gpio-header

For access to WebIOPi we have to enter your login and password. Default login is «webiopi», password - «raspberry». The browser will display WebIOPi interface, which to show all 40 pins of port GPIO and their destination. To change the login and password, we must enter the command:

sudo webiopi-passwd

For settings of WebIOPi under task we have to sensor pressure and temperature BMP180 register in the configuration file /etc/webiopi/config in section [DEVICES]:

bmp = BMP085

Figure 2 shows a wiring diagram of the sensor to pins of GPIO.

Fig.2. Connecting sensor BMP180 to the GPIO

In file /boot/config.txt we need to add a line: dtparam=i2c_arm=on

To change the password Webiopi we must enter the command:

$ sudo webiopi-passwd

Then we must restart the computer with help reboot command. To check the temperature sensor we need to connect to the address:

http://172.20.0.138:8000/app/devices-monitor

In browser, we should see temperature and pressure on the sensor (Figure 3).

Figure 3. The data is from the sensor BMP180

For overload WebIOPi after making changes to the configuration file, the Python script and an html file, you need to: /etc/init.d/webiopi restart

Error messages when you start Webiopi are in the file /var/log/webiopi. It can be printed on command: cat /var/log/webiopi

4. Creating file index.html and script on Python script.py

Figure 4. The index.html file

The need for these files is as follows. HTML-page via JavaScript makes the request to the script (the program), which written in Python. Python script returns the HTML-page the data, which obtained from the sensor BMP180 for their visualization. Every 5 minutes script records data about pressure and temperature in the text file. This file is used to build pressure and temperature graphs for the changing time. The contents of index.html shows in figure 4. It is in directory /home/pi/myproject/html . The content of file script.py on the Python shows on figure 5. It is written in directory of /home/pi/myproject/python

The file press.html shows in Fig. 6. Similarly looks temp.html file to generate a temperature graph.

Figure 5. File script.py

Figure 6. The file press.html for generate graph of pressure

For build graphs of pressure and temperature are used HTML files press.html and temp.html. These files use library the dygraph, which is written on the JavaScript. The file dygraph-combined-dev.js the library is copied from the site

http://dygraphs.com to directory /home/pi/myproject/html. After restarting the computer WebIOPi will work on the presented scripts. If you connect to it via a browser, information about pressure and temperature will be presented as shown in Figure 7.

Figure 7. Data from the sensor BMP180

If you click on the link "Graph pressure", in a new browser window will show a graph of pressure, similar to Figure 8.

Figure 8. Graph pressure is generated with help library dygraph

5. Connecting to the Internet of computer Raspberry Pi 2, if it has not got real ip-address or domain name, but has access to the Internet (via a modem, the router, firewall).

One way to get access to the Raspberry Pi as a device to the Internet of Things is the use Weaved service. It offers the following services:

- SSH - you can login in the Raspberry Pi from anywhere in the world via SSH;

- Web (http) on port 80 - you can view web - pages from anywhere in the world, located on the Raspberry Pi;

- WebIOPI - allows you to manage by pins GPIO Raspberry Pi, using software developed by the user.

Before installing the Weave you must to create the directory /home/pi/myproject/my, enter there and work there with Weaved files.

Installing Weaved on Raspberry Pi:

- You should get account on the site https://developer.weaved.com/portal/login.php;

- The Raspberry Pi 2 connect to the Internet;

- The Weaved Software download on the Raspberry Pi:

wget https://github.com/weaved/installer/raw/master/binaries/weaved-nixinstaller_1.2.13.bin

- File weave-nix installer v1.2.13.bin make executable:

chmod +x weaved-nixinstaller_1.2.13.bin

- Run the installation program:

./weaved-nixinstaller_1.2.13.bin

- Choose a service.

The first time you will be prompted to install one of the service:

SSH on port 22, Web (HTTP) on port 80, WebIOPi on port 8000, VNC on port 5091 (tested with tightvncserver), or a custom TCP on the selected port.

Select here the third service, Web (HTTP) in the 8000-th port.

- Enter your login information in the Weave (enter the account, which was received at Weaved site).

- Next, enter the name of your device, for example webiopi80.

- You must check, that was created a new device:

We go at https://developer.weaved.com/portal/login.php and enter your account.

After entering the next page (Fig. 9) should appear the name of the created device:

Figure 9. Listing created services

Conclusions.

1.Reliable remote control the sensors (equipments) with the help of mini computers via the Internet, compared with microcontrollers. On the microcontrollers the network protocols is lightweight, so do not work reliably.

2.The high cost of control systems on mini computers over the network compared to microcontrollers.

3.Thanks to software WeBIOPi we can simply program the mini computers that we use for remote management.

4. We are able to obtain access to the mini computers via the Internet in the event of inability to use the real IP - address and of the domain name. This is possible with the help of service Weaved.

5.The problem has a software module BMP085 when work with the pressure sensor BMP180. After a few hours of work is no longer work Web - WebIOPi server. Instead of is to work with the module BMP085:

from webiopi.devices.sensor.bmp085 import BMP085

bmp = BMP085()

better use module deviceInstance:

from webiopi import deviceInstance

bmp = webiopi.deviceInstance("bmp")

Here ("bmp") - is to get a device named bmp, which is in the file /etc/webiopi/config, in the section [DEVICES].

6. Mini computers can not only receive data from the sensors, but and process them.

Example - is plotting changes of values from the sensors.

Literature.

1. WebIOPi - The Raspberry Pi Internet of Things Framework. [Electronic resource]. - Mode of access: http://webiopi.trouch.com/, 2016.

2. Internet of Things for Everyone. [Electronic resource]. - Mode of access: https://www.weaved.com/ , 2016.

3. Комплексная система домашней автоматизации на Raspberry Pi. [Electronic resource]. - Mode of access: http://electromost.com/ , 2014.

4. Мясищев А.А. Интернет электро - розетка на основе мини компьютера Raspberry Pi и фреймворка WebIOPi. Практика для студентов. [Electronic resource]. - Mode of access: https://sites.google.com/site/webstm32/internet_rozetka, 2016.