Against the backdrop of advancing ecological civilization, strengthening ecological risk prevention in urban protected areas is of great scientific significance for improving conservation strategies and optimizing regional spatial planning. However, existing studies on ecological risk in urban nature reserves primarily rely on analyses of anthropogenic disturbance sources, lacking risk assessments directly linked to vegetation response data. Taking the ecological redline zones of Jiangsu Province as a case study, this research developed an integrated vegetation disturbance risk assessment framework that combines the Normalized Difference Fraction Index (NDFI), the Continuous Change Detection and Classification (CCDC) algorithm, and machine learning techniques. The goal was to achieve a direct and accurate evaluation of vegetation disturbance risk in urban protected areas. The results showed that: (1) the vegetation change detection achieved high accuracy, with an overall accuracy of 81.2% and a Kappa coefficient of 0.72, demonstrating that the CCDC algorithm effectively identified vegetation disturbances within Jiangsu’s ecological redline zones. (2) four machine learning models—binary logistic regression, random forest, extreme gradient boosting, and support vector machine—were tested, among which the random forest model performed best. The key driving factors of vegetation disturbance included river network density, slope, temperature, distance to city centers, population density, and gross domestic product, reflecting the combined effects of natural conditions and human activities. (3) vegetation disturbance risk exhibited distinct spatial differentiation, with high and very high risk areas mainly concentrated in ecological welfare forests and scenic areas where human–environment conflicts are prominent, while low and very low risk areas were widely distributed in key wetlands, important protection zones of Lake Taihu, and other ecologically stable regions. The findings provide a scientific basis for vegetation conservation in Jiangsu’s ecological redline zones and offer methodological insights for long-term vegetation disturbance monitoring and risk management in similar urban ecological areas.