Inversion Of Thermal Property Parameter Based On Improved Cuckoo Search Algorithm

The inversion of thermal property parameter has valuable theoretical importance and significant engineering value.The inverse problem can be solved by iteration methods or optimization algorithms.Optimization algorithms mainly include gradient algorithms and heuristic algorithms.The conjugate gradient method(CGM)has been employed as a delicate method,due to small storage space and computation time.The CGM is one of the local search algorithms and is sensitive to iterative initialization.As a new heuristic algorithm,the cuckoo search algorithm(CS)has a facilitating aptitude for the global search,which provides a powerful tool for solving inverse problems.But the CS has the property of slow convergence rate.Therefore,introducing the CGM into the CS is a feasible scheme for treating the inverse problems effectively and accurately.This approach is called as improved cuckoo search algorithm(ICS).The aim of this dissertation is to estimate thermal property parameter of two-dimensional heat conduction problems based on the boundary element method(BEM)and the ICS.For the estimation of thermal conductivity in nonlinear steady state heat conduction problems,Kirchhoff transformation is adopted to transform nonlinear heat conduction problems into linear heat conduction problems.For transient heat conduction problems without or with heat source,the inversion of thermal diffusion coefficient is studied.The heat conduction problem with heat source is transformed into the problem without heat source.The inversion process is presented for the estimation of thermal conductivity in nonlinear transient heat conduction problems,in which Kirchhoff transformation is also adopted.The numerical examples discuss the influences of element number,measured point number,nest number and measured noise,and so on.The result accuracy increases and the iteration number decreases with the increase of element number.The increase of measured point number causes the increase of iteration numbers,whereas the increase of nest number can decrease iteration numbers.Moreover,the greater the measured noise is,the lower the result accuracy is.The iteration numbers can decrease with the decrease of measured noise.Numerical examples show that the ICS is an accurate and efficient method for the inversion of thermal property parameter.