In order to cope with the large volume detection of the national rural road network and improve the low cost, convenience and operability of low-level rural detection, a low-cost and high-reliability prototype system for low-level rural road roughness detection is developed based 3-axis MEMS gyroscope, 3-wheel built-in MEMS accelerometer and dual GNSS navigation module. By analyzing the detection data of asphalt pavement and cement concrete pavement of more than 8,000 km rural highways, for low-grade rural highways, when the road condition is poor, a single flatness detection can be carried out to obtain the pavement performance evaluation results more reliably by mathematical model, and the detection strategy and scheme of large-scale road network detection of rural highways are proposed. RQI is used as a general detection index of rural highways. For sections with better road conditions, PCI sampling is carried out on the basis of RQI detection to improve the accuracy of pavement performance evaluation, so as to ensure good detection results and reduce the detection cost of rural highways.