A.A. Zhelnin

Dynamical errors in automatic control systems are generally caused by control principle « by mistake» and limitation of a pass-band the system. In automatic command missile guidance, as a rule, one count size of a dynamic error which consider at performing of commands of management. The first and the second derivatives of angular coordinates of the purpose are necessary for calculation of this error at loast. In conditions of placing the coordinator on a mobile chassis, the vising line acceleration and angular rate measurements on data from rate generator comes to appearance of noise, which are called by chassis vibration. It is possible to use rate generators info to form angular rate and line vising acceleration. To solve this mission is necessary to process it with low-pass filter fill maximum angular rate during traching the target on one place. After transforming into Cartesian coordinates it must be processed in differentiator. The main element of it is estimating coordinates that is showed in state. The main goal of filter adapting to chassis vibration condition is to calculate standart deviation of measurement error and it is compensation with varying filter indexes. Filter optimal indexes determination is easier to organize with finite quantify of sample, which must be cuffed by special sliding window. Besides, it is possible to input samples order decreasing during the removal from current time and to control the amount of accumulated samples, which are used to calculate measured altitude. At that index-math following concept provides filter characteristics optimization both for moving targets and coordinators ambiguity and for ambiguity in moving air target parameters measurement capability. As distinct of Calman-s filter, which are designed on the same filter indexes optimizations method, this digital filter is free from some minuses. Calman filter has essential fault: Calman-s filtration becomes insensitive to new educts. Method of controlling with such disadvantage lies in application of decreasing order of last measurement (exponential damping) whether removal from current measurement time. However, correct application of this method is connected with math solving of Riccati differential equation necessary. But this is not the case, and often obtained solving brings to divergence of filter, especially for air target differs from receiving one taken for solving supposal.