Sensor-Based Intelligent Adaptive Control for Plasma Process Manufacturing

该SBIR研究项目将开发并实现基于传感器的模型，将其应用于微波等离子体反应器系统上高密度、低压等离子体过程的智能、自适应和神经启发控制系统。目标是解决由微工厂/集群工具的传感器和组件的不确定性所产生的非线性和自适应控制方法的需求。为控制系统量身定制的特殊接口将使传感器、神经控制器和执行器的时间响应在1-5秒内，这足以实现实时和原位控制。此外，在实现算法时使用模拟VLSI芯片可将学习时间减少到最多几毫秒。这种计算需求的节省将确保有足够的时间进行精确和快速的操作。

This SBIR research project will develop and implement sensor-based models into an intelligent, adaptive, and neurally-inspired control system for a high-density, low-pressure plasma process on a microwave plasma reactor system. The goal is to address the need for nonlinear and adaptive control methods created by the uncertainty in sensors and components of the microfactory/clustor tools. Special interfaces tailored for the control system will render the time response of the sensors, the neuro-controller, and the actuators to within 1-5 seconds, which is fast enough to achieve real--time and in-situ control. Moreover, the use of analog VLSI chips in implementing the algorithms results in reducing the learning time to, at most, a few milliseconds. This savings in the computational requirement will ensure ample time for precise and fast actions.