In this paper a nonlinear and continuous model of Pulse Width Modulation (PWM)-based pneumatic position system with diaphragm-and-spring actuator is presented via theoretical analysis; against experimental data, the unknown parameters are identified by Particle Swarm Optimization (PSO). Using open-loop tests, the experimental data are obtained by controlling the on/off valves with three different duty ratios: 50, 80, and 100% over a period of 100 ms. The analytical-experimental comparisons show the validity of the model. The modeling is useful for intelligent valve positioner algorithm design, control valve optimization design and actuators fault diagnosis.