| Railway occupancy detection technology is one of the key technologies to ensure the safety of railway operation.The unilateral axle counting sensor has gradually developed into an important equipment in the field of railway occupancy detection with its simple structure,comprehensive function and high reliability,and has attracted the attention from more and more scholars.The sensor counts the axle according to the change of the induced electromotive force in the induction coils when the train passes.The induction electromotive force without wheel,the induction electromotive force with wheels and the change rate of induction electromotive force with wheels are important factors,which determine the performance of the sensor.Based on this,this dissertation take the unilateral axle counting sensor as the research object,studied the change law of induced electromotive force when the unilateral axle counting sensor parameters change by using magnetic circuit analytical method and magnetic field analysis method,and presented the parameter optimization method of the unilateral axle counting sensor.The analytical expressions of induced electromotive force are given based on the magnetic circuit analytical method,which will provide a guidance for the design and optimization of equipment.The axle counting principle of the unilateral axle counting sensor is analyzed.For the complex three-dimensional structure composed of sensor,rail and wheel with irregular geometric structures and their spatial position relations,the magnetic field segmentation method and the analytical method are combined to conduct the division of magnetic flux tube and the calculation of reluctance,and the equivalent magnetic circuit models are established at the two situations including without wheel and with wheels.According to the distribution characteristics of magnetic flux tube,the classification method of magnetic flux tube is proposed,the model of equivalent magnetic circuit is simplified,and the mathematical models of the sensor induced electromotive force are given based on magnetic circuit analytical method,and the analtical expressions of induced electromotive force are given at the two situations including without wheel and with wheels,and the effectiveness of the magnetic circuit analytical method is verified by simulation and actual measurementIn order to analyze the variation rule of the magnetic field near the sensor when the parameters change in detail,the calculation models of the distribution of magnetic induction intensity near the sensor in the three cases of without the rail and wheel,with rail and without wheel,with rail and with wheels are established,and the calculation method of induced electromotive force in the induction coils based on magnetic field analysis method is proposed.For the without rail and without wheel case,magnetic induction intensity calculation models inside and outside of the iron core are established based on the imaginary magnetic charge method.By studying the eddy current distribution in the rail,the calculation method of the magnetic induction intensity generated by the eddy current in the space near the rail is given according to the vector Green formula.By building a two-dimensional simplified physics model of a wheel and decomposing the eddy current field state of a moving wheel,the steadystate component model and transient component model of the wheel eddy current field are obtained.and the calculation models of the magnetic induction intensity generated by the eddy current in the wheel are established.The correctness of the proposed models is verified by simulation.In order to develop the unilateral axle counting sensor with excellent performance,an improved multi-population genetic algorithm is used to optimize the sensor parameters,and the optimal parameter matching to make the sensor with the optimal performance are given.According to the magnetic circuit analytical method and magnetic field analysis method,the key parameters that have great influence on the performance of the unilateral axle counting sensor are determined.The influence of each parameter changing on the induced electromotive force at the two situations including without wheel and with wheels is analyzed,and the functional relationship between each parameter and the induced electromotive force and the change rate of the induced electromotive force are given.A multi-objective optimization model is established for the change rate of induced electromotive force,the induced electromotive force without wheel and the induced electromotive force with wheels,and then the multi-objective optimization function is reconstructed to a single objective function by using the weight-composed method,and each weighted factor of the subobjective function is given through the judgement matrix method.A parameter optimization method of unilateral axle counting sensor based on improved multi-population adaptive genetic algorithm is established based on crossover rate and mutation rate adaptive method and large mutation method.The objective function values between the parameter matching selected by the optimization algorithm and other parameters matching are compared to verify the feasibility and effectiveness of the optimization method proposed through simulation calculation.Through the performance test of the optimized sensor,the correctness of the proposed optimization method is verified.The evaluation method of sensor performance is given by analyzing the change curve of induction electromotive force of induction coils when the horizontal distance between wheel and sensor induction coils changes.The experiment of sensor performance test is designed.Through high and low temperature test,electromagnetic compatibility test,vibration and shock test,the comprehensive performance improvements of the sensor after optimization under the conditions of without interference and with interference are verified,and then the effectiveness of the optimization method improving the comprehensive performance of the sensor is proved. |
PDF Full Text Request