| With the increase of car ownership in China,the situation of traffic safety has become more and more serious,traffic accidents have become the main cause of human casualties and road congestion,and more than 90%of car accidents are caused by human error.Driverless cars are considered to be the solution to modern intelligent transportation,which can improve traffic safety and reduce the burden on drivers.Path planning is the core of unmanned driving,and obstacle avoidance is the top priority of path planning.The artificial potential field method is widely used in path planning because it is simple and effective,but there are problems of minimum value and unattainable target,and the obstacle avoidance ability in dynamic environment is limited.According to the characteristics of the driving environment and obstacles of unmanned vehicles,this paper conducts the following research on the artificial potential field method:By thoroughly studying the classical artificial potential field method,clarifying its basic principle,performing the force analysis of the intelligent vehicle in the potential field,and discussing the shortcomings of the traditional artificial potential field method,i.e.,the minimum value problem of the algorithm itself and the obstacle avoidance ability of the traditional artificial potential field method in the dynamic environment,this paper optimizes the algorithm.The algorithm is optimised in this paper.Aiming at the problem of static single obstacle vehicle target unreachability,the gravitational potential energy is supplemented to the target point,and the distance fractional function is introduced into the repulsive potential field and gravitational potential field function to ensure the problem of proper energy replenishment,so that the vehicle can effectively escape the target unreachable trap.Aiming at the global minimum problem caused by unmanned vehicles,static obstacles and target points in the same straight line,a new virtual target point is introduced for the vehicle,and when the vehicle falls into the minimum trap,the location of the virtual target point is found,so that the vehicle reaches the virtual target point under the action of virtual gravity to escape the minimum trap,and then continues to move to the target point.Aiming at the minimum problem caused by the combined force of multiple static obstacles and the gravitational force of the vehicle target point,the increase function is introduced for the gravitational potential field function,and the gravitational potential energy of the target point on the intelligent car is dynamically increased to avoid the problem of insufficient gravitational increase near the minimum value and fall into the minimum point.For dynamic obstacles,the relative velocity and acceleration function of the unmanned vehicle and the obstacle is introduced into the repulsive potential field function to ensure that the vehicle moves away from the obstacle without considering the obstacle avoidance.When considering the road boundary,define the road repulsion potential field function,set up virtual obstacle points,and establish the road boundary repulsion potential field inversely proportional to the lateral distance from the vehicle to the left and right boundaries of the road,so that the intelligent vehicle can drive in the center area of the road without obstacle avoidance.The above five schemes have been simulated and studied to verify their effectiveness in the application of unmanned vehicle obstacle avoidance. |
