Research On Measurement Accuracy Compensation Of Ultrasonic Wind Speed Sensor For Mining

As the intelligent decision-making and emergency control of mine catastrophe ventilation continue to develop,there is a growing need for accurate measurement of cross-sectional wind speed in mine roadways.In the past,most mines relied on point wind speed measurements to estimate the average wind speed of the section,which often resulted in large measurement errors.However,with the emergence of ultrasonic wind speed sensors,it is now possible to measure linear wind speed,significantly reducing the error in estimating the average wind speed of the cross-section.Given that wind speed measurement in the mine is highly susceptible to the flow field and ambient temperature,this thesis proposes a novel approach that leverages the characteristics of ultrasonic wind speed sensors.Specifically,the authors first use Fluent software to simulate the relatively stable area of the flow field in the roadway,and determine the optimal location to install the wind speed sensor.They then utilize an improved Ant Lion algorithm-neural network(IALO-BPNN)to compensate for the temperature of the sensor,thereby enhancing the accuracy of wind speed measurement and reducing the deviation between the measured and actual wind speed values.To begin with,the article presents an overview of the measurement principle of ultrasonic wind velocity sensor.It then goes on to analyze the impact of laminar and turbulent flow on the flow rate of the flow field,and how this results in interference on sensor measurement.Additionally,the article examines how the temperature affects the characteristics of the ultrasonic transducer,which leads to deviations in the measurement time of the jet lag method.Secondly,leveraging the impact of flow field variables,the ANSYS Fluent software is employed to simulate the roadway and simulate the flow field distribution across varying pipe diameters.Through the establishment of diverse monitoring points,the stable zone of the flow field before and after the bend is determined,facilitating the strategic placement of sensors and mitigating the flow field’s interference with wind speed measurement.Furthermore,the experimental platform validates the flow field’s stability,specifically the region between 10 D after the bend and 4D before the bend.Thirdly,to address the impact of temperature on the ultrasonic wind speed sensor,BPNN is leveraged for compensation,while the Ant-Lion algorithm(ALO)is introduced to optimize the BPNN parameters.However,to overcome the limitations of ALO’s susceptibility to local extrema,three enhanced strategies are incorporated:adaptive guidance,Levy mutation operation,and simulated annealing.To validate the effectiveness of these improvements,the simulation of eight benchmarking functions is conducted.Consequently,the IALO-BPNN model is established as the temperature compensation solution.Lastly,an experimental platform is constructed for the ultrasonic wind speed sensor.Using the custom-designed sensor,a temperature change experiment is conducted,and the temperature compensation model,created using the sample data,is tested and trained.The results indicate that the use of the IALO-BPNN temperature compensation model yields an accuracy of 99.23% for the output wind speed value.The discrepancy between the output wind speed value and the actual wind speed value is a mere 0.09m/s,showcasing the exceptional precision of the system.This thesis includes a total of 56 figures,13 tables,and 92 cited references.