A method to optimize the structural parameters of the planarcoil linear time grating displacement sensor based on nonlinear programming genetic algorithm is proposed, which is used to improve efficiency and effectiveness comparing with the enumeration method. The nonlinear programming genetic algorithm is used to optimize the structural parameters of the sensor, and the sensor prototype based on optimized results is designed and experimented. Results show that the efficiency is enhanced by 65% and the harmonic distortion is reduced by 10 times, compared with experimental results using enumeration method. Meanwhile, in the range of 240mm, the measuring accuracy of the sensor can reach ±0.8 μm. This paper shows that the method restrains the original error induced by structure and the optimized sensor can realize high accuracy measurement.