Artificial Intelligence-Enhanced Peak Current Mode Control for Enhanced Power Regulation in High Power High Frequency Electrosurgical Generators

Abstract

This study presents an AI-enhanced peak current mode controller developed for electrosurgical generators (ESG). A regression neural network (RNN) is used by the controller to minimize peak current errors. This makes the generator more accurate and stable when there are non-idealities and varying tissue impedances. The controller’s efficacy is evaluated through rigorous MATLAB Simulink, processor-in-the-loop (PIL), and hardware results. The ESG is based on a buck converter and full bridge inverter. The MATLAB simulation results show good consistent output power, voltage, and current across several reference powers (30W, 40W, and 50W) and impedance values (i.e., 80 ohms, 100 ohms, and 120 ohms). The proposed controller has shown improved mean squared error (MSE) and root mean squared error (RMSE) values of 6.76 and 2.60 times, respectively, compared to previous control schemes. The proposed scheme is also verified by the real-time response of PIL simulations using a C2000 Launchpad. The prototype has validated the ESG performance on the C2000 Launchpad. This AI-powered control system has shown good performance in terms of accuracy, safety, and reliability, as verified through simulation and hardware results.

Keywords

• Artificial intelligence (AI)
• current mode control (CMC)
• electrosurgical analyzer (ESA)
• electrosurgical generators (ESG)
• neural network (NN)
• processor-in-loop (PIL)