Patrol Strategy Generation Algorithm And Application Based On Illegal Parking

In recent years,with the improvement of people's living standard and the change of people's living style,motor vehicles become the essential means of transportation in people's daily life.However,with the explosive growth of motor vehicles,parking place has become the focus of the people's attention.At the same time,this also led to more and more illegal parking incidents,which brings a lot of inconvenience to the traffic.Due to the limitations of patrol resources,it is very difficult for the management department to detect the occurrence of illegal parking in a limited resource.In this paper,we use the simulation data to combine Game Theory and Machine Learning for the first time,to deal with the situation of illegal parking incidents.In this article,according to the data of past years,we use the machine learning algorithm to find out the great attractive regional characteristics to the motor vehicles.And then we use gradient ascent algorithm to fit the optimal weighting parameters.The algorithm we used in this paper is an improved one,which is based on an existing algorithm.We increase the number of feature selection,to make the simulation more accurate.Then,we use the model based on Stackelberg model in Game Theory to build the patrol strategy generation model and use the optimal weighting parameters we got to calculate the optimal patrol mixed strategy.At the same time,we give the complexity analysis of the patrol strategy generation model and the error range of the approximation solution and exact solution.We can show the illegal parking information,vehicle information,owner information and the generated strategy information to the users through a set of B/S system.Meanwhile,the users can also update or insert or delete the data in real time.Among them,the illegal information table is the main table,in order to provide the necessary data for the algorithm.Vehicle information table and owner information table are the subordinate tables,these two tables do not provide the necessary data for the operation of the algorithm,just to provide users with some other information query for user reference.Finally,we gave an experiment which are comparing with other algorithms in this paper and we came to a conclusion: the algorithm in this paper is better than the other two algorithms on the efficiency and function payoffs.