Nonlinear Model Predictive Control and Collision-Cone-Based Missile Guidance Algorithm

In this paper, a hybrid missile guidance algorithm is proposed for planar engagement scenarios. The method combines nonlinear model predictive control (NMPC) with a collision cone approach, and enables the guidance algorithm to achieve interception, while ensuring that the impact angle lies in a predefined range. The guidance scheme comprises two components: 1) point-mass-based NMPC and 2) collision-cone-based NMPC. The point-mass-based NMPC is employed when the distance between the missile and the target is large, and when this distance falls below a threshold, the algorithm switches to a collision-cone-based NMPC. Connections between the impact angle and the collision cone are theoretically established, and these are used in the collision-cone-based NMPC. The NMPC problems are formulated as quadratic programs that include appropriate constraints on the states and inputs, while not requiring target acceleration information. These quadratic programs are shown to be strictly convex. Detailed simulations are performed to demonstrate the effectiveness of the proposed approach.