Research On Measurement Principle And Algorithm Influencing Factors Of Track Detection System

At present,the detection technology of track irregularity has been gradually perfected.There are quite mature detection methods both in foreign countries and in China.For the detection method of the inertial reference method in China,few articles summarize general methods for designing and implementing track detection systems,reveal the nature of track irregularity detection parameters,and analyze the impact of different track detection algorithms or operating conditions on the detection results.In this paper,the GJ-6 orbit detection system is used as a prototype,and the structure and main algorithm of the detection system are described in a template,and the two important core algorithm parts,namely the inertial reference method and the superelevation related factors,are discussed.The applicable conditions of the algorithm model of the current detection system are pointed out.The main research contents include the following aspects:（1）Each parameter of track irregularity is described as a function of mileage,and the connection between the chord measurement method and the inertial reference is established.Most of the descriptions of the various parameters of track irregularity in the traditional literature are separated.Based on the four important points of the transverse section of a certain track,namely the left and right profile points,and the positions of the left and right alignment points,this paper establishes the connection between these points and the mileage through the Frenet coordinate,and uses the function to describe the nature of the track irregularity parameters.The transfer relationship between after deriving the basic expressions of the chord measurement method and the inertial reference method can be easily obtained by using the established track irregularity function.（2）The general design method of the detection system based on the inertial reference method is summarized.In terms of hardware implementation,the necessary hardware including inertial component,laser displacement meter and encoder are described.In terms of software implementation,the necessary process for the system to output the track irregularity parameters is given.The filter technology used is described from the perspectives of the time domain and the space domain,and the relationship between the two is established through speed,and the fundamental reasons and general design methods for designing filters are explained.The above-mentioned concept of templated track detection system has important reference significance for rapid transplantation or design of a new system.Taking this as the minimized system model,all detection algorithms can be parameterized,thereby greatly improving the efficiency of testing and development of the detection system.（3）The influencing factors of the inertial reference method used by the existing system are analyzed.The ideal track irregularity is generated by simulation,and its corresponding ideal value of the detection sensor can be deduced.According to the algorithm flow of the detection system,the track irregularity parameters can be recovered.The error of the detection method can be obtained by comparing the generated value with the restored value.After adjusting some parameters and calculation methods,the main factors affecting the calculation results can be obtained.（4）A spatial domain complementary filtering method is proposed to calculate superelevation.The accuracy of the superelevation calculation is verified by simulating the curve and combining the actual test results and statistical test values.The related influencing factors of superelevation are also analyzed.According to the results of poor repeatability of superelevation in the actual detection process,it is found that the superelevation calculated by spatial domain filtering is more accurate.The main reason of distortion of superelevation detection when the speed changes too fast is analyzed.（5）The applicable conditions of the existing track detection system algorithm model are pointed out.By analyzing the level detection waveform of a high-speed comprehensive detection vehicle when there is a periodical shaking with a large amplitude and a long duration,It is found that under the existing sensor arrangement,the roll angle of the vehicle body relative to the ground level cannot be accurately measured.When the car body has a relatively large shaking motion,the roll angle of the car body relative to the track plane compensated and corrected by the displacement meter also has a large deviation.As a result,the repeatability of level,profile and TQI value become larger,so the accuracy of the measurement results of the detection system under the above working conditions is reduced.