On the determination of unspecified tf in a guided missile optimal control law application

The performance of any realistic optimal control law in a missile application is dependent on the estimation of final time, or equivalently, on time-to-go. Typically, an estimate of the range between the target and missile and the rate of change of this range is obtained from radar or other ranging devices; the time-to-go estimate is then calculated. This estimate works quite well as long as the range and range-rate information are accurate. In many instances, however, the data are contaminated by noise either covertly, as in the case of radar jamming devices, or by the processing electronics. This adversely impacts the estimate of time-to-go and the optimal control law; and missile performance suffers. This paper presents a discussion of several aspects of the problem in the context of a realistic application and provides analytic computer algorithms for its solution, as well as a closed-form result. An extension of the work is described in a recent publication [1].