Cellular Automata Technique For Predicting Failure Pattern Of Masonry Wall With Opening

Since Cellular Automata technique for predicting failure pattern of masonrywall is a discrete method, the predictions may have some drawbacks. For example,Some of the damage locations are made up of discontinuous points and some areconnected into a region. These cases are contrary to the real phenomenon ofmasonry wall damage and reduce the accuracy of the prediction. In order to improvethe precision and clarity of the prediction, we propose two solutions:First, we propose a matching criterion based on Taylor: We define thedifferences of the state value, the gradient and Heather Matrix on any position ofdigital modes between two walls. These differences are summed linearly to obtainthe total difference. A similar region is defined when the total difference is minimal.Then we use the new matching criterion and the matching criterion based onminimal error to predict the failure pattern of wall with opening. The results showthat the new matching criterion has significantly improved on fracture continuityand accuracy compared to the matching criterion based on minimal error.Second, we explore the reasons for discontinuous fractures in the failurepattern prediction from the perspective of noise, and propose two methods to reducenoise: Noise reduction of the difference function and the mapping relationship. Forthe difference function, we use the repeated noise reduction algorithm Instead ofGaussian filtering algorithm, in order to reduce the amount of computation; For themapping relationship, we establish a bilateral filtering algorithm of celldisplacement by simplifying the mathematical model. Examples show that thecontinuity of the failure pattern predictions using two noise reduction methods bothhave significantly improved.Subsequently, we discuss the computer technology associated with CellularAutomata technique for predicting failure pattern of wall:First, we establish a visualization platform of the wall fa ilure pattern predictionbased on Cellular Automata method. allowing the users to leave the code, they canachieve the failure pattern prediction of wall through interface.Second, a simplified model of wall cracking p rocess is proposed and anumerical simulation is achieved, thus providing a reliable basis for CellularAutomata technology.