优化方法用于模型校核

在工程实际中，常常会遇到需要选择本构模型和校准参数的模型校核问题。文中采用优化方法实现了复杂模型校核的自动化。文中，首先将各种本构模型映射为整数编号，并根据这些编号来激活相应的本构模型和本构模型参数，解决了本构模型的参数化问题；另外，根据模型校核问题的特点选择了试验结果和计算结果相对误差的平方和作为目标函数。通过这些步骤，用实现了优化方法模型校核自动化。

Model Identification with Optimization Method

There are many model identification needs in engineering practice, which commonly include constitutive model selection and parameter identification. And sometimes they involve the most complicated problem, such as turbulence calculation, which can only be solved with numerical methods. So, the traditional model identification methods used widely by control engineering are not fit for these kinds of problem whose model identifications are commonly accomplished manually. An optimization searching method was introduced to solve constitutive model selection and parameter identification. There are two problems during solution. First, how to select constitutive models automatically. Selection of constitutive model during calculation needs lots of settings and selection for simulation code and usually just can finish by hands function. This difficult was overcome by marking the constitutive models with numbers. And proper settings for each constitutive model can be selected on these mark numbers. Second, there is the difference of objective function between model identification and other optimization problems. For model identification, the objective function is to minimize the errors between test results and calculation values by selecting proper a constitutive model and model parameters. For this reason, the sum of squares of percentage errors was used as the objective function. By this way, the model identification problem was converted to an optimization problem. A model identification example of turbulence was provided by using the method mentioned above. And the result shows that this method can be used to replace the manual method for constitutive model selection and parameter identification in the complex problem.