Abstract:
This project designs a control system for a needle free jet injection. It first investigates the currently existing models for injection and then made some alterations based on basic theories in physics. These alterations yielded a new set of equations which were solved using the Runge-Kutta-Felberg formula. Second order, fourth order and fifth order curve fitting of the numerical model were performed in order to develop a novel polynomial that gives the direct non-differential pressure-time relationship for the needle free jet injection system. It was observed that the fifth order polynomial presented the least error, as measured by the least of the norm of residuals. The model for voice coil actuator used with the injection system was developed by writing force balance equations. The transfer function of the actuator was then used to investigate the behaviour of the injector in the skin. Finally, the two systems (voice coil motor and injector) were linked up by setting up an analogous electromechanical system to determine the behaviour of the voice coil motor in the injection system. The mechanical parts of this system represent the injection pressure profile with respect to skin dynamics. An adaptable PI (Proportional Integral) controller based on Model Response Adaptive Control (MRAC) using MIT (Massachusetts Institute of Technology) rule was designed for the full model to compensate for deficiencies arising from variations in skin properties for various individuals or species. The use of the adaptable PI control makes the work unique as it enable the device to deliver drugs to various skin types, location and layer just by calibration based on the trajectory as defined by the developed trajectory generator. At the end of this research work, the developed models were validated by comparison with other models from other researchers. The results shows adequate tracking of the adaptable system with the reference signal with a response time of less than 10ms which has a much better performance than other methods
