For the problem of the standard particle filter with slow response to the abrupt faults, the quantum-behaved particle swarm optimization (QPSO) is applied to the standard particle filter to diagnose gas path abrupt faults for gas turbine engine. A new conception about weight variation coefficient is introduced in the novel algorithm. Abrupt fault is detected and alarmed as weight variation coefficient exceeding a preset threshold. The QPSO algorithm incorporates the newest observations into sampling process to make the particles move to the high likelihood area, leading to more accurate estimates of parameters which change abruptly. Simulation results show that the improved algorithm makeｓ quicker response to the abrupt fault, compared with the standard particle filter.