About adaptive control of sequence of technological objects

We consider the problem of identification and control of a sequence of objects in conditions of incomplete information. Among the series-connected objects there may be objects both dynamic and inertialess with delay. This kind of process chains often takes place in the aerospace industry. They arise in the manufacture of rocket technology, electronic products, etc. The problem of integrated management of technological process group certainly belongs to the category of topical. This raises a whole range of problems of identification and control due to the fact that a priori information about them can be different. Below we consider the cases when the a priori information about the object has both parametric and non-parametric nature. We study the problem of identification in non-parametric or part non-parametric uncertainty. If a priori information is not sufficient for the initial statement of objectives, it is natural to use the theory of adaptive learning systems. The main purpose of the following paper is to give an algorithmic framework for modeling and process control within the production schedules. First, we consider the problem of identification and control for a local object, and then the task of managing a group of similar objects. Depending on the particular problem and a priori information level the corresponding modeling and control algorithms are designed. These algorithms are based on enough developed parametric and nonparametric theory of adaptive system. The theory of non-parametric systems is based on local approximation method, in particular nonparametric estimation algorithms of various kinds of dependencies on the results of the input-output variables of the object. The paper presents the specific nonparametric model of inertialess objects with delay and corresponding nonparametric control algorithms with memory. Control devices in this case are devices with memory, which makes nonparametric control algorithms with active accumulation of information the dual control algorithms.