Research On The Positioning And Navigation System Of Underground Parking Lot Based On WLAN

The rapid economic development and the improvement of people's living standards have led to the rapid growth of car ownership in the society.The underground parking lots of various social places,corporate units and residential communities have become increasingly popular.However,large-scale underground parking lots often have such problems such as complex terrain,unclear routes,hard to find spare parking spaces and difficulties in finding the car reversely.These problems seriously affect the user experience and the efficiency of using parking lots.The development of indoor positioning and navigation technology provides an effective way to solve these practical problems.Due to the limitation of the frequency bands used by GPS and other outdoor positioning and navigation systems,sufficient and effective signals can not be received in the indoor environment,and indoor location requires higher accuracy.Therefore,the traditional outdoor positioning method is difficult to extend to interior.As one of the signal mediums that can be used in indoor location technology,the Wi-Fi signal in the wireless local area network(WLAN)has good characteristics such as reliable stability,anti-interference ability and high data transmission speed.Using WLAN to achieve indoor positioning is completely based on the existing infrastructure.The function of sending and receiving Wi-Fi signals is already the basic function of all kinds of mobile intelligent terminals.Therefore,no additional special equipment is required,the cost is low and the layout is simple.Its accuracy can meet most requirements of location based service(LBS)and is therefore perfectly suitable for underground parking application of navigation and positioning.There are two main algorithms of the positioning technology based on WLAN: the geometric positioning method and the scene analysis method.Geometrical localization method has many unfavorable factors under indoor conditions.There are serious reflection,scattering and multipath effects when radio signals propagate indoors.Using single propagation model to calculate distances would have significant errors.The scene analysis method,which can be called the position fingerprint method,does not need to measure the distance and is not susceptible to the influence of the specific signal propagation mode,so it is more appropriate.This paper makes a positive improvement to the traditional position fingerprinting algorithm to improve the positioning accuracy and response speed as much as possible to meet the requirements of the actual system applications.In the off-line database establishment phase,the Gaussian filter method is used to preprocess the original data and then use the kernel C-means clustering algorithm to reasonably sort the database to reduce the complexity of the fingerprint matching and the amount of calculation,and due to the space-transform characteristics of clustering algorithm,can effectively eliminate the influence of the singular reference point.In the online positioning phase,the signal of the point to be positioned is preprocessed by Gaussian filter according to the distribution of the wireless signal.Then,the algorithm of fuzzy logic is used to replace the traditional fingerprint matching algorithm,and the cosine distance between the fingerprint vectors is introduced into the calculation of the similarity weighting coefficient which can significantly improve the positioning accuracy.Finally,the final coordinate of the point to be positioned is obtained by weighted K-nearest neighbor algorithm.The path in the parking lot is generally more regular driving channels,and After realizing the real-time location of the vehicle,the optimal parking path guidance and reverse vehicle search function can be realized by combining the path planning algorithm.After analyzing and comparing various path planning algorithms,this paper chooses more suitable Dijkstra algorithm,which has better global convergence,and can obtain higher optimal-path success rate.The main contents of this paper include:1.The relevant monographs and documents are summarized,and the existing indoor positioning technology is analyzed and compared,which provides the basis for the selection and optimization of this positioning scheme.2.Analyze the features of WLAN technology and its advantages for indoor positioning in parking environments.Summarize the content and basic process of location fingerprinting algorithm.3.The theories and algorithms involved in the Positioning navigation scheme of this paper are introduced,which includes Gauss filtering algorithm,kernel C-means clustering algorithm,fuzzy logic theory,Sugeno model and Dijkstra path-planning algorithm.4.The positioning and navigation technology scheme and algorithm design of this paper are described in detail.5.Demonstrate the method and result of the simulation experiment,verify the validity of the optimization algorithm adopted?6.Summarize the work of the whole paper,put forward the insufficiency of the current scheme and the direction of the next enrichment and optimization.