A real‐time quadrotor trajectory planning framework based on B‐spline and nonuniform kinodynamic search

Abstract

Autonomous navigation of quadrotor is required by many application scenarios, such as exploration, search, and rescue. The trajectory planning algorithm is the core of autonomous navigation, which can undoubtedly greatly enhance the safety of flight.In this paper, a trajectory planning framework based on B-spline and kinodynamic search is proposed. This framework can be used for a limited-sensing quadrotor to plan safe and dynamically feasible trajectories in unknown environments, and the flight is safe and effective along with these trajectories. First, a B-spline based nonuniform kinodynamic (BNUK) search algorithm is proposed to generate dynamically feasible trajectories efficiently. The characteristics of nonuniform search make the generated trajectories safe and reasonable time-allocation. Then, a trajectory optimization method based on control point optimization is proposed. The trajectory generated by BNUK is further optimized by solving a quadratically constrained quadratic programming problem. The smoothness of the trajectory is improved, and the control cost is reduced. Extensive analysis and comparative

Direct Bézier-Based Trajectory Planner for Improved Local Exploration of Unknown Environments

“…Fig. 2a reports the classical condition used for collision checking with Bézier curves [26], where the overall curve is constrained inside a safe sphere. The aforementioned approach often results in being too conservative, as a matter of fact the considered sphere is far to be tight over the convex hull, and thus over the curve itself.…”

Flatness-based Quadcopter Trajectory Planning and Tracking with Continuous-time Safety Guarantees

“…The safe flight space is usually non-convex which makes the obstacle avoidance problem difficult [14]. Obstacle avoidance with continuous-time safety guarantees was studied in [13] where each obstacle is expressed as a polytope and the trajectory planning problem is formulated as a mixed-integer QP, which is an NP-hard problem and renders the online trajectory re-planning computationally infeasible.…”

“…{victor.freiremelgizo,xiangru.xu}@wisc.edu formally safe trajectories in continuous-time [11], [12], [13], [14]. These works find a sufficiently smooth trajectory in the flat output space which can be mapped to the states and inputs of the system with algebraic transformations.…”

“…These works find a sufficiently smooth trajectory in the flat output space which can be mapped to the states and inputs of the system with algebraic transformations. Recent works along this line of research provide continuous-time safety guarantees in the form of state, thrust and angle constraints [11], [12] and obstacle avoidance [13], [14] constraints. The main challenge with the flatness-based method lies in the highly nonlinear mapping between state/input spaces and flat output space, which makes the consideration of safety constraints in the state/input space challenging.…”

Survey on Motion Planning for Multirotor Aerial Vehicles in Plan-Based Control Paradigm

“…Quadrotor attitude control for trajectory tracking [3], trajectory planning in large-scale and cluttered environments [4], a differentially flat hybrid system is used to formulate the trajectory planning problem for a bidirectional quadrotor [5], decentralized trajectory planning method for quadrotor swarm [6] Sampling-based method for time-optimal paths generation for a point-mass model [7], a continuous reference trajectory refinement technique for slow-speed maneuvering [8], trajectory planning approach considering geometrical configuration constraints and user-defined dynamic constraints for unconstrained control effort minimization [9], Logistic curve-based trajectory generation technique [10] Gaussian process-based residual dynamic learning [11], nonuniform kinodynamic search-based trajectory generation [12], a standard form of a two-point boundary-value problem using Pontryagin's minimum principle-based approach is proposed [13] Online teach and repeat planning technique was proposed [14], in which a geometric controller [15] was utilized for trajectory tracking. Moreover, an iterative trajectory refinement strategy was proposed to relieve the local minima problem where the free space was represented as a convex cluster, i.e., a set of convex polytopes [14], a faster approach for segmenting free space as a set of polytopes using point cloud [16], receding horizon trajectory generation was proposed in [17], whereas trajectory generation for moving target was proposed in [18] Trajectory planning technique was proposed based on nonuniform B-splines ensuring kinodynamic feasibility [19] where geometric tracking control (GTC) is used for controlling, incremental ESDF method for constructing the environment [20], B-spline based kinodynamic search algorithm followed by elasticbased optimization [21], preception-aware optimal trajectory generation with limited filed of view [22], direct collocation method for trajectory generation [23], Minimum-time B-spline trajectory generation [24] B-spline based kinodynamic search followed by refining the trajectory by using elastic optimization (EO) [25], fast marching method alone side with Bernstein basis polynomial trajectory generation [26], Topomap: three-dimensional topological map in which the sparse...…”