A Local Real-time Path Planning Method Based On Out-of-tolerance Of Flight

It is necessary condition for Unmanned aerial vehicles(UAV) to complete scheduled task chains, that flying along the desired trajectory accurately and reliably via automatic navigation. But in fact, UAV’s navigation system maybe failed or wrong for some unknown reasons, or other emergency tasks need to prioritize. All these can lead the scheduled task chains to be changed, make UAV difficult to complete it. This event called out-of-tolerance of flight, and the corresponding event information called out-of-tolerance information. In order to reduce the influence of scheduled task chains can’t be carried out smoothly. In addition to improve the reliability of UAV navigation system itself, accurate and timely monitoring and quickly processing to out-of-tolerance information is also very important. This article studies on this issue.This article introduces the concept of out-of-tolerance and it’s cause reasons first. And analyses the influence to scheduled task chains’ execution. And analyses the treatment methods of several typical kinds of out-of-tolerance in detail. Meanwhile, as the process of out-of-tolerance need local real-time path planning method, this article then analyses the local real-time path planning problem model, and discussed several kinds of algorithm’s applicability in the model. On the basis, this article points out the limitations of traditional algorithm such as genetic algorithm in dealing with local real-time path planning. then analyses the specific reasons of the limitations. And verify the analyses by experiment. All these give a direction for overcoming algorithms’ defects.To solve the problems that unable to plan back and request enough forward maneuver distance when path planning, this article puts forward a local real-time path plan method combined with orient-circle theory. As the similarity between orient-circle and continuous turning path. When the start point and end point’s direction has huge difference, the method can add a path segment in the two points’, to make the difference smaller, and avoid plan back problem. And when the two points are too close, this method can add an intermediate point apart from the two points. The intermediate point can guide the algorithm go through a roundabout way, and make the forward maneuver distance longer. Experiments show that the method can achieve a higher success rate when plan back is need or forward maneuver distance is not enough.This article’s method can overcome the problem of the traditional algorithms in local real-time path plan. And can dispose several typical out-of-tolerance quickly and reasonably. Experiments results prove the method’s correctness and validity.