China has a vast number of cardiovascular disease patients, and blood pressure monitoring is essential for prevention and management. Traditional cuff-based devices are inconvenient and unsuitable for continuous use, while cuffless methods using PPG and ECG signals offer an innovative alternative. This project develops a systematic framework for cuffless blood pressure monitoring, including data collection, preprocessing, quality evaluation, and prediction. A CNN-based PPG quality index model is combined with a VAE-based unsupervised learning method to build a multi-level quality assurance system capable of detecting abnormal signals. At its core, an improved 1D V-Net model with multi-channel design integrates ECG, PPG waveforms, and pulse arrival time to achieve precise blood pressure prediction. The system improves the accuracy of noninvasive monitoring and provides reliable support for cardiovascular disease prevention and management, with broad clinical application potential.