| 摘要: | 本計畫〝腦波導引自走式看護系統之系統電路設計與實現〞主要目的在於製作腦波導引自走式看護系統之腦波擷取/放大電路、無線腦波傳輸器、機械手臂/自走車馬達控制與驅動電路、硬體浮點運算單元以及即時HHT(希爾伯黃轉換法)訊號演算法硬體實現等電路。首先，本子計劃欲開發腦波擷取/放大電路用以取代現有的昂貴腦波量測儀器。為了方便使用者活動以及搭配控制自走車的移動，本計劃另設計一無線腦波傳輸器用以將收集到的腦電波訊號傳送至控制端進行腦波訊號辨識。馬達驅動電路設計部分是在於機械手臂/自走車之控制與驅動上的相關設計。本計畫中，直流伺服馬達將被會配合受測者腦電波進行控制轉動，是屬於一種變結構控制系統。本計畫之計畫主持人在變結構控制領域已有多年的研究亦有相當的成果，並將之應用在機械手臂控制上。針對機械臂/爪控制，預計推導出適用機械肢體系統之變結構控制方法。在演算法硬體實現部分，預計以FPGA (場可程式邏輯閘陣列)為基礎來實現演算法的硬體電路，並使用VHDL(硬體描述語言)來設計FPGA 硬體電路。有別於一般所使用的以DSP(數位訊號處理晶片)為核心做設計，本計畫演算法硬體實現部分是採用全 FPGA 純硬體方式作設計，執行演算法算數運算速度將可大幅提升，實現即時的訊號處理能力。本大腦人機界面系統可以用來讓肌肉損傷、漸凍人、大腦運動神經損傷與重度癱瘓等病人，能夠利用腦波快速且正確的控制自走車與機械臂的抓取功能與外在環境溝通。 This proposal, “System circuit design and implementation of brain signal guiding autonomous nursing system”, is a part of the whole proposal, “Design and Implementation of brain signal guiding autonomous nursing system”. The object of this proposal is to design the control algorithms for motor drive, and to implement the signal processing algorithm –Hilbert-Hunag Transform (HHT) on the FPGA. In the first place, the proposed system, different from the BCI system using bulky and expensive electroencephalography (EEG) measurement equipments, personal computer, and commercially real-time signal processing software. The proposed system combines a customized stimulation panel, a brainwave acquisition circuit, and an FPGA-based real-time signal processor and allows users to use their brainwave to communicate with or control multimedia devices by themselves. At the second part of the proposal, the motor driver circuit design is about the control and drive of the mechanical joint motor. In this proposal, we can’t just use the traditional control theory to drive the motor. Because the system is belong to variable structure system. The host of the proposal has many research results in the variable structure systems field in many years, and apply the control theory on the robot control. We will base on the research result to development the control method for the proposed system. At the third part of the proposal, algorithm implementation, we will use the FPGA (Field Programmable Gate Array)-based to implement the ICA algorithm. And the design will use the VHDL (VHSIC Hardware Description Language) to construct the FPGA hardware. In order to speed up the hardware algorithm, this proposal will use full-hardware FPGA design. The speeds of mathematics operation will faster then traditional design using DSP-based. Therefore, it will achieve the real-time signal process and it can control the powered exoskeleton system in time. The proposal is divided into three-year project. The main tasks for each year are listing at following: First year: brainwave acquisition and wireless brain wave transceiver circuit design and implementation, robot arm control and drive. Second year: module of float-point and HHT algorithm hardware implementation on FPGA, and module synchronization design Third year: pipeline hardware HHT architecture design and implementation, and ADC/DAC related interface circuit design. 研究期間：9908 ~ 10007 |
