Study On Multisensor Road Spectral Acquisition System And Its Roughness Algorithm

With the rapid development of the transportation industry,the number of uneven roads has increased dramatically,which greatly affects the driving comfort,reduces the performance and service life of the roads.How to timely get road information accurately and real-time detect the roughness of uneven roads effectively has become a hot issue of concern.Road roughness detection technology and uneven road detection algorithm are designed to solve these problems and have become a hot topic in the field of intelligent transportation.The paper makes a deep research on uneven road detection system and road roughness detection algorithm based on the research of uneven road data characteristics and the principle of road roughness detection.The works in this paper are as follows:1.This paper constructs the road roughness detection system which is composed of acceleration sensor,laser displacement sensor and rotary encoder: In order to develop the road roughness detection system of data acquisition part and realize the multi sensor acquisition measurement function of accelerometer and laser displacement combination,this paper chooses Freescale ARM Cortex-M4 core K60 chip as the main controller based on analyzing the system feasibility.According to the study of the multi sensor technology,the detection of uneven road is realized through developing the connection between acceleration and laser displacement sensor.2.In this paper,the road reasoning rules is compiled and the hierarchy diagram is designed according to the overall structure: The Protégé platform and Jena API tools are selected to compile road reasoning rules after analysing the reasoning and query function of the system,so that the intelligent reasoning between the ontology model and the road information can be developed.Then the SPARQL statement is used to achieve the road information query based on semantic web.3.In order to improve the accuracy of road roughness detection detection,this paper proposes the road roughness detection algorithm based on optimized RBM deep neural network: The inputs of algorithm are the vehicle vertical acceleration power spectrum and the pitch acceleration power spectrum,then Adaboost_Bp algorithm is used in each optimized RBM deep neural network classification model for fine-tuning due to its performance of global searching.The collected data are analyzed by optimized RBM deep neural network and the experimental results show that the system has a good detection performance and achieve the expected effect.