Mathematical model and algorithm of accumulation of information by the robot with non-absolute memory

Formulas are given that determine the accumulation of information by robots that have non-absolute memory and are psychological analogues of man. It is shown that, with constant information memory coefficients of such robots, the accumulated information is limited. The formulas allowing under these conditions to determine the values of the accumulated information by the robot to a given time point, called the information cycle, and the maximum value of the accumulated information are given. The definition of anomalous accumulated information corresponding to the infinite accumulation of information by a robot is introduced. Theorems describing the properties of accumulated information and allowing for some types of variable information memory coefficients to eliminate the restriction on the amount of accumulated information with simultaneous partial forgetting are proved. An example of a function describing the change in the coefficients of information memory with simultaneous infinite growth of accumulated information is given. A dimensionless characteristic of the accumulated information is introduced, allowing to determine the speed of the anomalous information accumulated by the robot at a specific moment in time. To illustrate the behavior of the function of information memory coefficients and the accumulation of information, the corresponding graphical dependencies are given. The ways of using robot theory with non-absolute memory in the functioning of call-center robots are proposed. An algorithm for accumulating information about clients by a call center robot operator providing an individual approach to each client is proposed, and a method that allows to regulate the behavior of the dimensionless characteristic of the accumulated information about the client.