On The Safety Control And Distribution Optimization For Multi-Autonomous Vehicle System

Multi-autonomous vehicle system has been a hot research topic in the intelligenttransportation field in recent years because it can release complex workload to people,andprovide timely service for traffic network.In this paper, the main research work is the safetycontrol strategy in some typical roads,the configuration optimization and task allocationmethod in distributed environment,which is based on the control and optimizationstrategy.The details are as follows:Firstly, the multi-autonomous vehicle system model is established and a filter isdesigned to estimate the autonomous vehicle motion state. The autonomous vehiclelongitudinal dynamics model, handling dynamics and multi-autonomous vehicle systemsmodel,is established respectively.Then the analysis of the impact of air resistance on theautonomous vehicle longitudinal speed, and the analysis of the reasonableness ofautonomous vehicle design and corresponding stability under different frequencies forhandling dynamics are made. The collaborative modes of multi-autonomous vehicle systemsare described combined with the hybrid automata model. The Kalman filter is designed toestimate the autonomous vehicle motion state.The simulation results demonstrate therationality of the model.Secondly,the comfortable following strategy and the flexible,efficient intersectioncollaboration stategy are studied.Considering the emergency brake for vehicle damage,thesafety following problem is settled by introducing the natural exponential function. For theT-shaped intersection safety problem, a method of safety coefficient estimation is proposedto estimate the desired force and the vehicle speed,combined with an idea of artificialpotential field.An increment-type digital PI controller is used to achieve the accurate controlof the vehicles.The simulation results verify the safety following control strategy and thecollision avoidance planning of the autonomous vehicles in the T-shaped intersection basedon safety coefficient estimation are practical and effective.Thirdly,the configuration optimization and task allocation to meet a desired averagecoverage and reduce the average cost of time are settled.A” static to dynamic” idea isused.At first,the simulated annealing algorithm is used to solve the the coverage respect todifferent numbers of autonomous vehicles.On this basis,the configuration number ofautonomous vehicles meeting the specified coverage is obtained relatively easily in thedynamic case.For the task location problem,a structure of decision-making module based onthe BDI model is given,and the Hungary algorithm is utilized to obtain the task allocationscheme.The simulation results demonstrate the feasiblility of the configuration optimization by coverage statistics,the validity of the task allocation by the proportion of the averagearrival,completion rate estimates and average time cost.Finally,a multi-autonomous vehicle system structure is given and the system simulationplatform is built to analyze the performance of the system modules.The indoor positioningsubsystem module is tested for its accuracy. The positioning perfermance is analyzed with areasonable set of parameters.The software platform for experiments conducts networkingcommunications subsystem to realize locaion information transmission and control of theautonomous vehicle.In summary,a series of theoretical studies about safety control and distributionoptimization strategies are conducted in this paper.The main purpose is to realize the safetycontrol for the autonomous vehicle and distribution optimization for multi-autonomousvehicle system by the corresponding verification and analysis in simulation experiment.Thesimulation platform is built to test the system performance.