Research On The Optimal Patrol Strategy Of Urban Management Based On Bounded Rationality

In the streets of Chinese cities,some illegal pedlars are often seen selling some fake and inferior products and unhygienic food to people who do not know about this,which may cause serious health problem.In addition,the process of illegal pedlars illegally selling in the streets of the city will produce a lot of garbage,which affects the environmental sanitation of the city.Besides,they often cause people to gather and so may lead to traffic accidents.To address this problem,the urban management department needs to send urban inspectors to patrol the urban area,so as to manage these peddlers' illegal selling problem,maintain the urban environmental hygiene and ensure the safety of the citizens.However,for a large area of urban areas,the urban inspectors' patrolling resources are,after all,limited,and it is difficult to patrol all areas of the city.Therefore,under the condition of limited patrolling resources,the problem of how to maximize the management of illegal selling of some peddlers has become an urgent problem for the city management department.Thus,urban management department needs to allocate patrol resources according to the characteristics of the illegal pedlers' behaviors,so as to solve the pedlers' illegal sale problem more effectively.The resource allocation method based on Stackelberg security game provides an effective theoretical model for the problem.Therefore,on the basis of the Stackelberg security game,this paper proposes a patrolling resource allocation method based on the bounded rational adversary model,TALENTED(Target Attributes LEarNing model with TEmporal Dependence).This method can effectively analyze the adversary' behavior characteristics,predict the adversary' behavior trend,and help the city management department to generate better patrol strategy.TALENTED provides two main contributions.First,a new adversary behavior model is proposed to predict the probability of each target(streets of Chinese cities)being attacked,and the new adversary behavior model consists of three aspects:(1)This model takes into account more environmental characteristics that can affect the decision of the adversary's behavior,and fully considers the effect of the target environment on the adversary's behavior;(2)As the adversaries want to maintain their longterm benefits,they tend to be at one place.Therefore,this model takes into account the impact of adversaries' previous behaviors on their future decisions.(3)Compared with the adversary behavior model in the previous literature,in the new adversary behavior model proposed in this paper,the environmental characteristics that can affect adversary's behavior decision has different parameter vectors,which can predict the adversary's behavior more accurately.Second,this paper adopts the resource allocation method based on the Stackelberg security game to calculate the optimal patrol strategy for the urban inspectors.Simulation results illustrate the TALENTED method can help urban management create a reasonable and efficient patrol strategy and effectively solve the pedlers' illegal sale problems.