International Journal of Applied Engineering Research ISSN 0973-4562 Volume 9, Number 6 (2014) pp. 665-673 © Research India Publications http://www.ripublication.com
Application of Fuzzy Logic for Temperature Control in Microcontroller Based 2.45 GHz Microwave Hyperthermia Device Imam Santoso1, Thomas Sri Widodo2, Adhi Susanto2, Maesadjie Tjokronagoro3 1
Doctor Candidate of Electrical Eng., Dept. of Electrical Eng. & Information Tech. 3 Faculty of Medicine, Gadjah Mada University, Yogyakarta, INDONESIA E-mail:
[email protected] 1, 2
Abstract In this paper we describe the implementation of fuzzy algorithm embedded in microcontroller for temperature control in our developed microwave hyperthermia device. As the standard of hyperthermia therapy (using microwave heating mechanism) keeping the temperature of radiated diseased object must be at 41-45oC in a certain heating time. We use pulse width sequence with duty cycle variation derive by a few fuzzy rules to control the temperature of heated object or bio-medium. In this experiment an artificial biological medium representing the human body tissue was used. As the hyperthermia experiment result our microwave hyperthermia system had shown the temperature stability in specified hyperthermia temperature during microwave heating. Keywords: fuzzy; hyperthermia
microcontroller;
temperature
control;
microwave
1. Introduction Hyperthermia is a process of heating the tumor tissue in human body (surface or inside the body) in the therapeutic temperature range from 41-45oC. This temperature range must be maintained as long as treatment duration
666
Imam Santoso et al
(approximately 30-60 minutes) in order to have safety heating for healthy tissue surrounds the tumor [7]. In microwave hyperthermia device, hardware point of view, microcontroller can be used to control the microwave tissue heating [1, 2, 3, 10], with the support of invasive [2, 3, 9, 11] or noninvasive [8] (noncontact) thermal sensor measured the object surface or near surface temperature to be heated. There are many control algorithms can be implemented in the microcontroller, such as simple on-off method [2, 3, 10], classic PID (Proportional Integrator and Differentiator) [1], fuzzy [6], or hybrid fuzzy-PID. Because of the nonlinearity response of the biological tissue, fuzzy and fuzzy-PID are more suitable for nonlinear environment, but in this research we choose the simple fuzzy for simplicity the microcontroller programming, other to investigate that simple fuzzy algorithm can also have stable response as the advance one.
2. Microwave Heat Transfer and Fuzzy Logic Control As describe above microwave hyperthermia system is a kind of temperature control system of biological tissue that have been heated by microwave radiation with finite duration time, tH. It behaves like a nonlinear plant according to the heat transfer equation in bio-medium (1). ρC
T
ĸ
T
Q
Q
Q
1
ρ is bio-tissue density, C is bio-tissue specific heat, ĸ is bio-tissue heat conductivity, T is tissue temperature, Qb is heat transfer due to blood perfusion, Qm is metabolic heat transfer, Qe is external heat transfer from microwave radiation source, Qm and Qb can be ignored in our ex vivo experiment as in [4, 10, 11], and using pulse width signal to control the duration of microwave heating then (1) become ρC
T
ĸ
T
Q w t
2
w(t) is pulse periodic signal (the period = tw) with duty cycle variation in time, according to duty cycle function Δ(t), refer to Fig. 1. The pulse duty cycle influence the tissue temperature response, wider duty cycle, longer microwave expose to bio-tissue, and cause more heating time. A is pulse amplitude indicated on state and 0 for off state of microwave source, and n is 0, 1, 2, 3… N-1, (N = tH/tw), tH is hyperthermia duration. This correspond to typical unit step response of a control system, the elimination of errors can be realized using combination of longer Δ(t) (reduce positive errors) and smaller Δ(t) (reduce negative errors) of the pulse sequences.
Application of Fuzzy Logic for Temperature Control
667
Figure 1: Pulse sequences with constant duty cycle (a) and with variable duty cycle (b)
The generation of w(t) can be realized using microcontroller unit (MCU), it can be automatically turned on or off the magnetron (as microwave source) by some kind of algorithm. Here we choose fuzzy logic algorithm to control the Δ(t) in w(t) in order to keep stable the temperature response of nonlinear microwave hyperthermia system. The application of fuzzy logic as control algorithm, have been widely use for controlling many non-linear devices. It is not just easy to code and implement, other is that the need of knowledge precision of the plant model or parameters is unnecessary, the operation is basically depend on heuristic logic decision, if-then relation. Beside the advantages, surely the careful design of the fuzzy logic must be highly concerned.
Material and Methods Hardware Design The radiated type of microwave hyperthermia device that has been build comprises (refer to Fig. 2 and 3) of a magnetron (M) (from National YJ1530SP) as a 2.45 GHz microwave source or generator with output power maximum 310 watts, then the microwave feed into a semi conical aluminum applicator (A) via microwave flexible coaxial cable (C) as a waveguide.
668
Imam Santoso et al
Figure 2: Block diagram of Microcontroller Based 2.45 GHz Microwave Hyperthermia.
Figure 3: Our developed 2.45 GHz Microwave Hyperthermia System (left) and its MCU schematic diagram inside box. The control signal is generated by microcontroller unit (MCU) depend on temperature value taken from a thermal sensor a noncontact thermopile array
Application of Fuzzy Logic for Temperature Control
669
(S), it drives the magnetron to be operated in a pulse width power mode, this mode can eliminate the error in temperature reading, continuous microwave heating could influence the thermopile sensor, so as the switched-off state of magnetron the temperature value is taken by the sensor. This microwave hyperthermia experiment use a medium sized box-shaped agar phantom representing ex vivo biological tissue or medium.
Fuzzy Logic Control Design The fuzzy logic control (FLC) diagram as describe in Fig. 4, has 2 inputs membership function error and Δerror and one output membership function, this output responsible for generating pulse width signal (PWS) w(t) from MCU, the corresponding duty cycle Δ(t) due to on state of microwave heating are 0%, 10%, 25%, 50%, 75%, and 90% (Fig. 7). These Δ(t)s decisions are derived from 49 fuzzy if-then rules (Table 1). We note that UH= largest or maximum pulse duration with duty cycle Δ(t) 90%, HH= upper middle pulse duration with Δ(t) 75%, LH= middle pulse duration with Δ(t) 50%, SH= lower middle pulse duration with Δ(t) 25%, CC= smallest pulse duration with Δ(t) 10%, all duty cycles Δ(t) taken from pulse periode tw (refer to Fig. 2). The error e and Δe both have the same fuzzy membership function, there are PB for positive big, PM for positive medium, PS for positive small, ZZ for zero, NS for negative small, NM for negative medium, and NB for negative big.
a
b
670
Imam Santoso et al
c Figure 4: Fuzzy input membership function for (a) Temperature Error (b) Δ Error (c) Fuzzy output membership function (γ) for Duty Cycle Variations.
The fuzzy rules R* relate between e and Δe, using general relation if e and Δe then γ. All of these fuzzy membership functions, fuzzy rules, and defuzification algorithm was coded in C++ and then mounted in a microcontroller unit (MCU) AVR ATmega32. Table 1. Fuzzy Logic Rules for Duty Cycle variation R* PB e PB HH PM HH PS HH ZZ HH NS CC NM CC NB CC
PM HH HH HH HH CC CC CC
PS HH HH UH UH SH CC CC
Δe ZZ HH UH UH LH SH SH CC
NS NM NB UH LH SH LH LH SH LH SH SH SH SH CC SH SH CC SH CC CC CC CC CC
Result and Discussion The temperature responses of microwave hyperthermia system with or without fuzzy logic control condition have shown in Fig. 5. In open loop (Fig. 5 left), uncontrolled condition, or using continuous power it can be seen that temperature raise far above hyperthermia temperature, cause burn effect at the surface of the heated bio-tissue object. Temperature response of microwave hyperthermia system with FLC (Fig. 5 right) shows the transient state (about 10 minutes) and the steady state after 10 minutes. The FLC embedded in MCU shows the ability to control microwave heating according to given hyperthermia temperature set point
Application of Fuzzy Logic for Temperature Control
671
(44oC) using pulse width signals with its appropriate duty cycle variation, for the fuzzy output decision refer to Fig. 6.
Figure 5: Hyperthermia temperature response of open loop (uncontrolled) (left) and fuzzy controlled (right) microwave heating time 1500 seconds.
Figure 6: The corresponding pulse width signal sequences generated by MCU
Conclusion We have developed a fuzzy implemented microcontroller based 2.45 GHz
672
Imam Santoso et al
microwave hyperthermia device and as the experiment result this device has a stable temperature response as long as hyperthermia heating duration.
Acknowledgment The author thanks to Indonesian Directorate of Higher Education for scholarship fund, and also Electronic Lab. of Electrical Eng. Dept., Diponegoro University, Semarang, Indonesia for microcontroller software and hardware assisted.
References [1] R. Knochel, W. Meyer, F. Zywietz, Dynamic ‘In Vivo’ Performance of Temperature Controlled Local Microwave Hyperthermia at 2.45 GHz, IEEE Microwave Theory Techniques Society International Microwave Symposium Digest, 1982, pp. 444-447. [2] R. L. Magin, T. S. Fu, R. E. Beard, C. A. Cain. Local Tumor Hyperthermia Using a Computer-controlled Microwave System. Bioelectromag., Vol 3, Issue 3, 1982, pp 363–370, . [3] P Nilsson and B Persson. Computer controlled microwave system for clinical hyperthermia. Phys. Med. & Biol. Vol. 30 Issue 4, 1985 (DOI:10.1088/0031-9155/30/4/001) [4] Wesley L. Nyborg, Solutions of The Bio-heat Transfer Equation, Phys. Med. Biol., Vol. 33, No. 7, 1988, pp. 785-792, U.K. [5] ARIB Standard T-56, Specific Absorption rate (SAR) Estimation for Cellular Phone, Association of Radio Industries and Businesses, January 27, 1998, Japan. [6] T. Ding, H. Ying, S. Shao, 1999, Structure and Stability analysis of Takagi Sugeno Fuzzy PI Controller with application to tissue hyperthermia therapy, Soft Computing 2, Springer Verlag, 1999, pp 183190. [7] Szasz, A.; Szasz, O.; Szasz, N. Physical Background and Technical Realizations of Hyperthermia. Landes Bioscience and Springer Science, 2006, pp. 27-59, . [8] Thomas Sri Widodo. Microwave Hyperthermia System Development Based on Non Invasive Temperature Monitoring. Proc. of Int. Conf. on Electrical Eng. & Informatics, ITB, Bandung, Indonesia, 2007, pp (E-18) 461-463.
Application of Fuzzy Logic for Temperature Control
673
[9] N.M. Roman, V. Dan, R.V. Ciupa, V. Pompas. Hyperthermia Control Using a Computer Microwave System in Cancer Therapy, Proc. of IFMBE 25/VI, pp. 103-105, 2009, Munich. [10] Vasile Surducan, Emanoil Surducan, Radu Ciupa, Marius N. Roman, Embedded System Controlling Microwave Generators in Hyperthermia and Diathermy Medical Devices, Proc. of IEEE Int. Conf. on Automation, Qual. & Testing, Robotics, 2010, pp.366-371, Cluj-Napoca, Romania. [11] Teng Jiao, Hua Wang, Yang Zhang, Xiao Yu, Huijun Xue, Hao Lu, Xijing Jing, Hua Zhan, Jianqi Wang, A Coaxial-Slot Antenna for Invasive Microwave Hyperthermia Therapy, Jurnal Biomedical Science and Eng. 5, 2012, pp. 198-202 (DOI:10.4236/jbise.2012.54026)
674
Imam Santoso et al
