With the continuous deepening of electricity market reform and the increasing proportion of renewable energy integration, the volatility of nodal electricity prices has become significantly stronger. How to quantify the impact of price perturbations on regional load has become a key issue for achieving flexible control on the demand side and formulating differentiated pricing strategies. This paper, focusing on the response mechanism of regional nodal price signals, proposes a three-stage progressive analysis framework for quantitative research on price perturbations. First, high-frequency historical load and price data are used to estimate regional demand price elasticity, and price response sensitivity levels are classified. Second, an interpretable machine learning method (SHAP) is introduced to analyze the marginal contribution of price volatility features in load forecasting. Third, multiple-perturbation scenarios of varying magnitudes are constructed, and an XGBoost forecasting model is used to simulate dynamic load responses under price perturbations, establishing a nodal price–load response association ratio model. An empirical study takes the US PJM market as an example; results show that loads in different regions exhibit significant heterogeneity in response to price perturbations, with high-elasticity regions responding more sensitively. The magnitude and frequency of price perturbations have nonlinear impacts on load fluctuations. The research outcomes can provide quantitative support for designing differentiated pricing mechanisms, deploying demand response resources, and optimizing ancillary services.