The Predicted Mean Vote (PMV) model, widely used in indoor thermal comfort assessment, performs poorly in low-pressure and non-uniform thermal environments. Existing modified PMV models are limited by their reliance on uniform conditions and standard atmospheric parameters, while traditional models for evaluating thermal sensation in non-uniform environments involve complex calculations. This study innovatively integrates the Local Thermal Sensation (LTS) assessment model with the pressure-corrected PMVC model to propose a novel PMV correction method. This approach effectively addresses the challenge of coupling low pressure with environmental non-uniformity, enabling rapid thermal comfort prediction based solely on input environmental parameters. The proposed method was applied to a typical low-pressure, non-uniform thermal environment scenario-the aircraft cockpit. Results indicate that the most thermally sensitive regions for cockpit occupants are the head, hands, feet, and calves. Notably, solar radiation significantly influences local thermal variations, with thermal sensation differences of up to 0.62 units observed between pilots' heads. Based on these sensitive regions, the study develops localized PMVL calculation formulas for 16 body parts and determines an overall thermal comfort evaluation formula: O-PMVL = 0.7829PMVC - 0.3802, incorporating the weight of each body part's contribution to overall thermal comfort. The findings significantly expand the applicability of the PMV model, enabling its use in aerospace and extreme environments. Additionally, the study provides key indicators and guidance for optimizing thermal environment design, facilitating advancements in thermal comfort design for various industries operating in specialized environments.