TSR Algorithm And Experimental Study On Ultrasonic Gas Temperature Profile Reconstruction

The measurement of the temperature distribution in a boiler plays a crucial role in the online monitoring of the combustion process, the optimization of the combustion, the reduction of the pollutant emission and the safe operation of the equipments. Common temperature distribution measurement methods can be divided into two categories, the invasive and non-invasive methods. The invasive methods benefit from the simplicity of the equipments and easy implementation. However, the temperature distribution of the measurement objects may be affected by the insertion of the probes due to the invasive measurement nature, leading to measurement errors. The non-invasive methods do not disturb the temperature distribution of the measurement objects, and overcome the drawbacks of the invasive methods, which has attracted increasing attention. Owing to the distinct advantages, such as the mechanical wave characteristics, the good adaptability to the specific conditions, including the poor lighting conditions, the high-concentration dust, the poisonous and harmful gases, and the strong electromagnetic radiation environment, the ultrasonic method is considered to be a promising temperature distribution measurement approach. This dissertation studies the ultrasonic temperature measurement method.The main contributions of the thesis can be summarized as follows.(1) According to the wave equations and the thermodynamic equations, the ultrasonic measurement principles are clarified, the relationship between the ultrasonic velocity and the media temperature is established, and the importance of the TOF data to the accurate measurement of the temperature distribution measurement is revealed.(2) The ultrasonic temperature distribution reconstruction procedure is an ill-posed problem, in which the accuracy of the reconstruction algorithm plays a vital role in practical applications. Different from common reconstruction methods, in this study a two-stage reconstruction (TSR) method that considers the inaccurate properties of the measurement data and the reconstruction model and the dynamic evolution information of the measurement objects is proposed to improve the reconstruction quality. In the first stage, in order to alleviate the ill-posedness of the ultrasonic inverse problem, a coarse grids is first used to discretize the measurement domain to ensure the over-determined property of the reconstruction model, and then an appropriate algorithm is employed to compute the temperature distribution. In the second stage, based on the radial basis function method and the temperature distribution data from the coarse grid system, a new objective functional that simultaneously considers the TOF data error, the inaccuracy of the reconstruction model and the dynamic evolution information of the measurement objects is proposed to cast the temperature reconstruction task as an optimization problem. The simulated annealing method is employed to search for the optimal solution of the proposed objective functional. Numerical simulation results indicate that the proposed reconstruction method can improve the reconstruction quality. As a result, a proposing method is introduced for the ultrasonic inverse problem.(3) An ultrasonic temperature measurement system, including the single chip microcomputer, the ultrasonic transducer, the amplifying circuit, the multi-way switch, the phase comparator, the schmitt trigger and other peripheral circuits, is design to collect the TOF data. The theoretical model and the Monte Carlo method are employed to analyze the errors of the hardware system. Under the considerations of the influences from the electronic devices and the environment, an appropriate range of the error for the ultrasonic transducer is assumed, which provides the scientific supports for the achievement of the high-quality temperature distribution measurement and the selection of the system hardware.(4) The experimental studies are carried out to validate the feasibility of the phase difference method, the amplitude phase modulation method and the threshold method in measuring the TOF data. Finally the threshold method is employed to measure the TOF data. At the same time, a new method and a practical measurement procedure are developed to compute the number of the wave length that is a challenging problem in the ultrasonic single frequency phase difference method. The feasibility of the method is experimentally validated.(5) With the consideration of the inaccurate property of the TOF data, a wavelet miltiscale analysis based method is proposed for the de-noising processing of the TOF data, which can improve the quality of the TOF data and establish a good foundation for the high-quality reconstruction of the temperature distribution.(6) The designed ultrasonic temperature distribution measurement system is used to measure the TOF data in a gas medium heated by a heat source, thereby obtaining the temperature distribution. The experimental results validate the feasibility and effectiveness of the designed ultrasonic system and the proposed reconstruction method. As a result, a promising method is introduced for the temperature distribution measurement in a gas medium.