SBIR-STTR Award

Modern computation techniques such as parallel processing are well-suited for the solution of such problems as the simultaneous global extremization of multiple variables in a complex, nonlinear, probabilistic state space. However, the use of these techniques to solve a given problem requires that it first be formulated in these terms. Currently, threat-avoidance mission planning is approached as a dynamic programming problem, and not as a global extremization problem. Thus, in order to apply modern computing techniques to the real-time determination of evasive maneuvers for an aircraft, or for any other weapons platform, the problem must first be redefined as a global, multi-dimensional extremization problem. We will use the technique of functional integrals to reformulate the threat-avoidance mission planning problem as a global extremization problem in a multi-dimensional, probabilistic state space. The consequences of choosing different threat functions will be evaluated and we will determine the optimal form of the threat function. We will evaluate the operational utility of this approach and we will predict the optimal computation technique for its implementation.Anticipated benefits/potential applications:The problem of threat avoidance is a generic one and the results of this project can be directly applied to weapons platforms other than missiles. For example, the problem of collision avoidance of aircraft is one area where the results can be directly implemented. Also, the analysis can be inverted to determine the optimal guidance algorithm for missiles.