This paper presents an Artificial Intelligence (AI) integrated novel approach to Brain-Computer Interface (BCI)-based wheelchair development, utilizing a voluntary Right Left Hand Movement mechanism for control. The system is designed to simulate wheelchair navigation based on voluntary right and left-hand movements using electroencephalogram (EEG) data. A pre-filtered dataset, obtained from an open-source EEG repository, was segmented into arrays of 19x200 to capture the onset of hand movements. The data was acquired at a sampling frequency 200Hz in the laboratory experiment. The system integrates a Tkinter-based interface for simulating wheelchair movements, offering users a functional and intuitive control system. Various machine learning models, including Support Vector Machines (SVM), XGBoost, random forest, and a Bi-directional Long Short-Term Memory (Bi-LSTM) attention-based model, were developed. The random forest model obtained 79% accuracy. Great performance was seen on the Logistic Regression model which outperforms other models with 92% accuracy and 91% accuracy on the Multi-Layer Perceptron (MLP) model. The Bi-LSTM attention-based model achieved a mean accuracy of 86% through cross-validation, showcasing the potential of attention mechanisms in BCI applications.

翻译：本文提出一种集成人工智能（AI）的脑机接口（BCI）轮椅开发新方法，利用左右手自主运动机制进行控制。该系统基于脑电图（EEG）数据，通过模拟左右手自主运动来实现轮椅导航控制。研究采用开源EEG数据库中的预滤波数据集，将其分割为19×200的数组以捕捉手部运动起始信号。实验室实验中的数据采集采样频率为200Hz。系统集成基于Tkinter的交互界面用于模拟轮椅运动，为用户提供功能完善且直观的控制系统。研究开发了多种机器学习模型，包括支持向量机（SVM）、XGBoost、随机森林以及基于注意力的双向长短期记忆网络（Bi-LSTM）模型。随机森林模型获得79%的准确率。逻辑回归模型表现出优异性能，以92%的准确率超越其他模型，而多层感知机（MLP）模型准确率达到91%。基于注意力的Bi-LSTM模型通过交叉验证获得86%的平均准确率，展现了注意力机制在BCI应用中的潜力。