热传导反问题智能化识别

Identification of the Inverse Heat Conduction Problem with Intelligent Optimization Method

基于一种时域精细算法和蚁群算法,利用测量信息和计算信息构造最小二乘函数,将多宗量反演识别问题转化为一个优化问题,建立了求解多宗量一维瞬态非线性热传导反问题的智能优化数学模型。可对非线性内热源强度、导温系数和边界条件等多个热学参数进行组合识别。对信息测量误差作了初步探讨,数值验证给出令人满意的结果。结果表明该计算模型能够对非线性多宗量热传导反问题进行有效的求解,并具有较高的计算精度。

A general intelligent optimization numerical model is presented to identify multi-variables of the nonlinear one dimensional inverse heat conduction problem in transient state by a precise algorithm for direct heat conduction, based on Ant Colony Algorithm. The inverse problem is formulated implicitly as an optimization problem with the cost functional of squared residues between calculated and measured quantities. Combined identifications can be carried out for non-linear source term, thermal diffusivity and boundary conditions etc. Satisfactory numerical validation is given including a preliminary investigation of effect of noise data on the results. Results show that the proposed numerical model can identify combined parameters for the inverse heat conduction problems with precision.