基于改进云推理算法的塔架结构损伤识别

为了解决塔架结构的损伤识别问题,提出了基于应变能和改进云推理算法的损伤识别方法。首先描述了云模型的基本理论和数字特征,并给出了模态应变能的基本公式;然后分析了X条件云发生器和Y条件云发生器的基本算法和运行步骤,借助灰云模型建立相应的前件云和后件云规则,考虑了测量噪声的影响,利用云发生器生成多组云滴,并利用多模式下云滴的确定度和生成值构建了基本云推理算法及其损伤识别指标。基本云推理算法中常会产生不均匀发散的云滴,从而使计算结果产生一定的偏差,为了降低云滴发散产生的偏差影响,提出了基于损伤模式数量加权的云推理改进策略。计算结果表明:云推理算法可以较好地应用于塔架结构的损伤识别,其识别结果明显优于传统的应变能耗散率指标方法;而改进云推理算法进一步提高了识别的精度,优于基本云推理算法。

Damage identification of tower structure based on improved cloud reasoning algorithm

In order to solve the damage identification problem of tower structure,a damage identification method based on modal strain energy and improved cloud reasoning algorithm is presented.Firstly,the basic theory and numerical characteristics of cloud model are introduced,and modal strain energy formulas are described.Then,the basic algorithm and calculation steps of X and Y condition cloud generators are analyzed.The qualitative rules of forward and backward cloud are established by using grey cloud model.Considering the effect of measurement noise,plenty of cloud droplets are generated by using cloud generators,and a basic cloud reasoning algorithm based on certainty and cloud value is presented.In general,the generated cloud droplets by the basic cloud reasoning algorithm are divergent,which will result in some calculation errors.In order to reduce the effect of divergent cloud droplets,an improved weighting strategy based on damage patterns is proposed.Simulation results show that the cloud reasoning algorithm can identify the damage of tower structures and the identification results of the algorithm are better than those of the traditional modal strain energy dissipation ratio index.The improved cloud reasoning algorithm improves the identification accuracy and it is also superior to the basic cloud reasoning algorithm.