Robustness Analysis Method Based On Reverse Model And Its Application In Tolerance Design Of Acoustic Metamaterials

Uncertainty in the manufacturing process of engineering products cannot be avoided.Usually,excessive machining error will lead to significant performance degradation of the product,or even the product cannot be used.Otherwise,when the processing requirements are too strict,it will bring expensive manufacturing costs.Especially for the engineering products with high parameter dimension and high precision requirements such as acoustic metamaterials,it is the key to ensure the usability and manufacturability of products to determine the maximum acceptable machining error within the allowable performance variation,which could balance the relationship between manufacturing cost and product performance.Adopting the "reverse model",it can analyze the robustness of the existing design scheme from the acceptable objective variation and constraint variation.It is an effective way to determine the maximum tolerance of uncertain parameters.In this paper,the robustness analysis method of engineering products is taken as the research object.According to the "reverse model",the robustness of the existing design scheme is analyzed by utilizing the inner and outer double nested framework,and the maximum tolerance interval of uncertain parameters can be solved.On this basis,taking acoustic metamaterials as a typical application scenario,the classification model is combined with the robustness analysis method based on "reverse model" to reduce the simulation cost.Finally,through the robustness analysis of the acoustic metamaterial design scheme,the maximum robust tolerance of each parameter is obtained under different performance requirements.The specific research work is as follows:(1)This paper proposes a robust analysis method based on the "reverse model".The proposed method maps the acceptable objectives variation region and acceptable constraints variation region given by the designer to the parameter-space.In the sensitivity region of the parameter space,by analyzing the robustness of the worst case variation,the maximum tolerance interval of uncertain parameters which satisfies the objective robustness and constraint robustness is obtained.The proposed method adopts the framework of inner and outer double nested,in which the outer loop searches the maximum tolerance interval,and the inner loop searches the worst case variation for robustness evaluation.Two numerical examples are used to verify the effectiveness of the proposed method under the assumption of unilateral acceptable objectives variation and bilateral acceptable objectives variation.The results show that the proposed method can obtain the maximum robust interval of uncertain parameters under design requirements.(2)In order to reduce the expensive simulation cost of acoustic metamaterial problem,this paper introduces SVM classification model and combines it with robustness analysis method based on "reverse model",which greatly saves the simulation cost.On this basis,the relationship between the performance requirements of acoustic metamaterials and the allowable machining errors of parameters is analyzed.By using the proposed robustness analysis method,the corresponding maximum tolerance intervals of parameters are obtained under different acceptable performance requirements.