A low-cost solution for controlling human body vibrations.

Lache, Simona ; Luculescu, Marius Cristian ; Barbu, Daniela Mariana 等

1. INTRODUCTION

Within the medical rehabilitation a specific category is represented by the persons who have suffered neural-motor injuries. In order to improve the motor functions or to diminish the symptoms of the disease, the orthesis are used for quite a long time by now.

The present work is part of the research project ID_147, financed by the Romanian National Council for Scientific Research in Higher Education and developed at Transilvania University of Brasov. The project aims to develop and implement an intelligent orthesis for the rehabilitation of the inferior/ superior articulations of the persons suffering of neural-motor problems. The training of the injured member is done by the information received from the healthy member, through a command and control unit (Krebs, H.I. et. al., 2003). The block diagram is presented in Figure 1. The device is attached to the injured arm/ leg, in the region the rehabilitation is required and it may be programmed in terms of the desired task.

One of the main roles of such an orthesis is to give the chance to the person with disability to have a life close as much as possible to the normal one. This can be achieved by the advantages such a solution offers: low cost, easy to adapt for personalized features.

A special category identified is that of the people working in a vibration environment that induce hand-arm vibrations. In this case, the training process from the healthy arm to the injured one has to take into consideration the additional "noise" induced by vibrations. This can affect the command and control system stability, so that identifying solutions for reducing the level of vibration would be of much interest. There are already some very well known solutions adopted for reducing the vibrations from the hand-arm system: the use of anti-vibe gloves or anti-vibe systems attached to the vibrating equipment. However, they do not cover the problem; for example, when working with vibrating tools of small dimensions, when neither the gloves or the anti-vibe systems are of any use, due to the precision required and the small size tools (i.e. dental technicians are exposed to hand-arm vibrations while working with various appliances and tools).

Ones the vibrating source is identified, the first strategic step for reducing its level is to use the most suitable damping solutions (Mansfield, N.J., 2005). The work presented in this paper proposes a data acquisition system for controlling the vibrations induced in the human body, as part of the overall design process for the orthesis device presented in Figure 1.

[FIGURE 1 OMITTED]

2. PROBLEM FORMULATION

Vibration signals can be usually acquired using professional systems containing a specific hardware and a notebook, having high prices. We will try to solve this problem by developing an own data acquisition system that has to fulfill the following conditions: low-cost, microcontroller-based and stand-alone system so that to control data acquisition process without the need of a computer. For doing this we will use a microcontroller that has an embedded analog-to-digital converter (ADC), the system containing also a keyboard and a liquid crystal display (LCD), for setting acquisition parameters, starting and stopping the process and displaying the evolution of the signals in time and different other results.

This paper focuses on the design part of the system suitable for such type of applications.

3. THE MICROCONTROLLER DATA ACQUISITION SYSTEM

The structure of the data acquisition system is presented in Figure 2. For designing the data acquisition system the most important component, the microcontroller, is considered in the first place (Heidemann, J., 2004). There is a wide offer in this field, important families of microcontrollers being offered by Microchip Technology, Atmel Corporation, Intel, Philips Semiconductors, Infineon and so on.

An 80C552 ROMless single-chip 8-bit microcontroller was chosen, manufactured in an advanced CMOS process and being derivative of the 80C51 microcontroller family, having the same instruction set as the 80C51 (Balan, 2001). It can also be chosen the other two derivative circuits, 83C552 with 8 kbytes mask programmable ROM or 87C552 with 8 kbytes EPROM, but 80C552 is cheaper. The microcontroller has an 8-bit data bus and a 16-bit address bus, allowing the use of 64KB of external memory. It contains 256 bytes of internal read/write data memory, five 8-bit I/O ports, one 8-bit input port, two 16-bit timer/event counters (identical to the timers of the 80C51), an additional 16-bit timer coupled to capture and compare latches, a 15-source, two-priority-level, nested interrupt structure, an 8-input ADC with 10-bit resolution, a dual digital-to-analog converter (DAC) pulse width modulated interface, two serial interfaces (UART and I2C-bus), a "watchdog" timer and on-chip oscillator and timing circuits (http://www.nxp.com, 2002).

[FIGURE 2 OMITTED]

This microcontroller has to satisfy all the conditions imposed by our application. It contains an 8 channels analog multiplexer allowing acquiring information from 8 analog signal sources.

The ADC is a 10-bit resolution one, so a sample will need 2 bytes for storing. The maximum quantity of external memory that can be addressed using a 16-bit address bus is 64KB. Due to the fact that the used microcontroller is ROMless, 4KB will be implemented with an EPROM module storing the system programs, the other 60KB being used for data storing in the acquisition process. Using 60KB of SRAM memory, a number of 30K samples can be stored. If the interested domain for the vibration signals is 0-200Hz, this means we have to used a sampling rate of minimum 400Hz, taking into account the Shannon sampling theorem. The 400 samples acquired in a second suppose 800 bytes for storing them, in the case we want to store data locally. So the 60KB of memory will be enough for 76.8 seconds of continuous recording at a sampling rate of 400Hz. If we need longer interval, a special external storing system has to be developed.

A conversion cycle takes 50 machine cycles, this meaning about 50 [micro]is for a system clock frequency of 11.0592 MHz, so the maximum sampling rate for this ADC can be about 20 KHz.

Taking into account all of the above, an 80C552 microcontroller satisfies the conditions imposed by the application.

The inputs of the analog multiplexer and of the ADC accept voltages in the range of 0 / +5V ([AV.sub.REF-] = 0V, [AV.sub.REF+] = 5V), supplied by the block for adapting signals.

The conversion result can be computed by relation (1).

N = [2.sup.10] x [AV.sub.IN] - [AV.sub.REF-] / [AV.sub.REF+] - [AV.sub.REF-], (1)

where [AV.sub.IN] is the analog input voltage and [AV.sub.REF-], [AV.sub.REF+] are the microcontroller reference voltages (Figure 3).

Special function registers ADCON and ADCH are used for controlling the acquisition process. ADCON selects the analog channel for the multiplexer (ADR0, ADR1, ADR2), starts the conversion (ADCS=1), announces the end of conversion (ADCI=1) and contains the less significant bits of the result (D1 and D0). ADCH contain the most significant part of the result (D9 to D2). For controlling the acquisition process, the system is provided with a keyboard and a graphic LCD with a resolution of 128x64. The keyboard contains the following 16 keys: 0 to 9 digits, Menu for activating menu mode, Enter for validating a setting or entering in a menu option, Cancel for aborting a setting, Up and Down for browsing the menu options, Start for starting the acquisition process. Data can be transferred to PC through an UART interface at baud rates between 300 and 115200.

[FIGURE 3 OMITTED]

4. CONCLUSIONS AND FUTURE WORK

The system acts similar to a datalogger, but it is a low-cost one. It contains a large number of options making it a very flexible tool in a wide range of applications that need data acquisition. Future developments will take into account:

* Using an 87C552 microcontroller instead of 80C552, due to the 8 KB EPROM for storing system programs, having in this mode all 64KB of external memory for data storage;

* Attaching an external storage system (USB accessible) for large data quantities;

* Sending data through Ethernet using a serial-to-ethernet adapter, wired or wireless;

* Using microcontrollers with large address bus and higher resolution.

The system proposed in this paper may represent a first step in creating an alternative to expensive professional data acquisition systems, being capable to solve simple problems, having a lot of advantages at a low-cost.

5. REFERENCES

Balan, R. (2001). Adaptive Control Using 80C552 Microcontroller, in Acta Technica Napocensis-Construction Machines. Materials No. 1 pp. 55-60, ISSN 1224-9106

Heidemann, J.; Govindan, R. (2004). An Overview of Embedded Sensor Networks, In: Handbook of Networked and Embedded Control Systems, D. Hristu-Varsakelis and W.S. Levine, editors, Springer-Verlag, 2004

Krebs, H.I., et. all (2003). Robotic application in neuromotor rehabilitation, In Robotica 21, No.1, pp. 3-12

Mansfield, N.J. (2005). Human Response to Vibrations, CRC Press, ISBN 0-415-28239-X, Boca Raton, Florida

***(2002). 80C552/83C552 Single-chip 8-bit microcontroller datasheet, http://www.nxp.com/acrobat_download/ datasheets/80C552_83C552_4.pdf, Accessed on 2009-04-15