Business Process Optimization Method Based On Petri Net Mining And Configuration

Nowadays,with the development of electronic information technology,it is more convenient and facilitated for users to withdraw money from ATM of Banks.Users in the withdrawal process of various operations will leave a lot of withdrawal log.Through arrangement and analysis of the saved log,it is effective to establish a more perfect business process model and at the same time is effective to the bank's control of the ATM,improve the implementation efficiency of the withdrawal business,guarantee the reliability of the withdrawal,and guarantee the security of the withdrawal,which has important practical significance.In this dissertation,Petri net and behavior profile are used to mine the process model from the log,and the model is configured and optimized by finding the change region and using the algorithm of three transition optimization models to make the model more reasonable.The main research of this dissertation is as follows:(1)Considering the establishment of model in target business process,this dissertation proposes a method of mining business process from event log.This dissertation based on weak order relation of Petri net behavior profile to analyze the behavior relation between activities in a log,proposes an algorithm to mine business process from the event log based on Petri net,which firstly extract logs to classify and count the number of the traces,select more frequent traces to analyze behavior profile,find the relationship between each traces,establish the initial model before using the evaluation criteria of the model's rationality value to calculate and analyze the model,then adjust model to finally determine the optimal model.(2)Considering the configuration problem of the change region in the business process,this dissertation proposes a method of configuration optimization after finding the change domain.The existing determination of the change region only adds the control structure from the structure,generating control structures being difficult,and there is no effective algorithm to configure the changing part.In this dissertation,two algorithms are firstly proposed to find and determine the variable domain.Firstly,an algorithm for finding sets of variable points in a target model is used to find the set of variable points in the target model.The algorithm outputs the set of variable points in the target model by inputting a target model,calculating the set of variable points before and after the set of boundary points,and the transition relation of these point sets.Secondly,an algorithm for determining change region in the target model is used to determine the change region of the target model.This algorithm takes the variable points in the first algorithm as input and outputs the change region of the model by calculating the compliance and support of the sub-model.Then,a Petri net configuration optimization algorithm is proposed for configuration optimization of Petri nets.This algorithm takes the set of change region in the model as input,and outputs the change region of the configured model after allow,hidden and block transitions are configured.Finally,this dissertation specifically explains the finding of the change region and the change of configuration optimization by considering the problem that different users have withdrawal limit in ATM.Figure[18]table[8]reference[70]...