In this study, a stochastic simulation-optimization model (SSOM) is developed for planning electric power systems (EPS) under uncertainty. SSOM integrates techniques of support-vector-regression (SVR), Monte Carlo simulation, and inexact chance-constrained programming (ICP) into a general framework. SVR coupled Monte Carlo technique is used to predict the electricity consumption amount; ICP is effective for reflecting the reliability of satisfying (or risk of violating) system constraints under uncertainty. The SSOM can not only predict the electricity demand exactly, but also allows uncertainties presented as interval values and probability distributions. The developed SSOM is applied to a real-case study of planning the EPS of Shanghai, with an objective of minimizing system cost and under constraints of resources availability and environmental regulations. Different scenarios associated with SO2-emission policies are analyzed. Results are valuable for (a) facilitating predicting electricity demand, and generating useful solutions including the optimal strategies regarding energy sources allocation, electricity conversion technologies, and capacity expansion schemes, (b) resolving of conflicts and interactions among economic cost, electricity generation pattern, SO2-emission mitigation, and system reliability, and (c) identifying strategies for improving air quality in Shanghai through analyzing the economic and environmental implications associated with SO2-emission reduction policies. (C) 2014 Elsevier Ltd. All rights reserved.