Research On The Key Technologies Of Pedestrian Navigation Considering Characteristics Of Sensors And Pedestrians' Needs

Walking is one of the most important modes of travel in today's city life.When pedestrian walks to a destination in an unfamiliar environment,the pedestrian navigation service can help the pedestrian to complete the path finding task,which is of great significance to people's travel life.The popularity of smartphones has promoted the development of pedestrian navigation services.Nowadays pedestrians always carry smartphones with them wherever they go.The multiple types of sensors the smartphone carries can then provide a variety of data related to pedestrians and the environment.These phones can function as service platforms as well as data acquisition devices.Pedestrian navigation service research is based on people.Pedestrians have the ability to autonomously recognize the surrounding geographical environment and make decisions in the navigation process.Therefore,analyzing in depth the needs of pedestrians in different navigation environments and stages is of great importance to improving navigation service's quality and comfort.According to the existing pedestrian navigation service,the entire process of pedestrian navigation service can be divided into three key stages:positioning,route planning and route guidance.Based on the sensor data of the smart phone platform and the very basic idea of focusing on the needs of pedestrians,this paper conducts research and discussion related to the three key phases of pedestrian navigation.The main research contents of the paper are as follows:1)A multi-source sensor data fusion positioning method based on evidence theory is proposed.In this part,smartphone sensors are used to collect multi-source data related to the surrounding environment.Then based on the evidence theory,the sensor features are taken as location evidence to represent the spatial location of paths.The synthesis rules of evidence theory are improved,and weighted processing is performed according to sensor characteristics to promote accuracy.Finally,the evidence synthesis method is used to complete the fusion of multi-source sensor data and realize positioning method.Experiments show that this method can realize sensor data fusion positioning and can better meet pedestrians' navigation and positioning requirements.2)A multi-objective route planning algorithm for nighttime environment is proposed.In this part,the space-time characteristics of the physical environment are considered.There is still a lack of relevant research on the special nature of the night environment in the existing navigation research.In view of the particularity of the nighttime urban environment,from the comprehensive analysis of urban crowds'nighttime activities and pedestrian psychology,two key indicators of environmental brightness and pedestrian flow were extracted.A multi-objective path planning model based on ambient light brightness,pedestrian flow and path length was constructed.Artificial bee colony algorithm is used to solve the multi-objective programming problem.Experiments show that the algorithm can effectively solve multi-objective path planning problems.3)A pedestrian guidance method using Google Tango's relative spatial awareness technology was proposed.Based on the nature of pedestrian's relative spatial perception,the route guide description experiment is used to test the individual route guidance preferences and abilities.The SBSOD experiment is used to evaluate the pedestrian's orientation and location perception ability.The ability is refined to five dimensions of data.Then conduct the analysis between pedestrian's perception of orientation and route guidance preference.By using Google Tango's human-like relative spatial perception capabilities,we can use the known trajectories to select environmental landmarks to achieve differentiated path navigation guidance.