Research on parameter optimization algorithms for acoustic collaborative measurement of multiple physical fields

Abstract

In the combustion process of large boilers, in order to ensure sufficient and stable combustion of combustion materials, it is necessary to monitor the physical fields such as temperature and flue gas flow fields during coal powder combustion in real time. Considering the refraction of sound waves, acoustic collaborative reconstruction model for multiple physical fields based on radial basis function in the furnace is established. The improved Tikhonov regularization algorithm is used to obtain a new regularization matrix and solve the ill posed problem. Optimization research was conducted on various parameters that affect the quality of physical field reconstruction in collaborative measurement processes. The shape parameters of the radial basis function were optimized using the equilibrium optimization algorithm, and the regularization parameters in the Tikhonov algorithm were determined using the robust generalized cross validation method. The reconstruction steps for the acoustic temperature field and velocity field collaborative measurement were provided by simultaneously optimizing the two parameters. To verify the feasibility of the proposed parameter optimization algorithm, numerical simulations were conducted to reconstruct the typical temperature field and tangential velocity field in the furnace using multiple physical fields. In addition, the combustion process of high-temperature swirling flow is simulated and non-uniform temperature and complex flow fields are simultaneously reconstructed in a furnace. In end, an experimental acoustic measurement system is developed to measure the temperature and velocity fields and evaluate our proposed method. All the results confirm the feasibility, effectiveness and noise immunity of our proposed method to simultaneously reconstruct multi-physics fields under different complicated environment.