a)The aim of the project is to create a small monoblock device for long-term collection of information on the state of the cardiovascular system with mathematical processing of the collected information for the last six episodes of recurrence in patients with atrial fibrillation. b)The author’s method of measuring the change in blood pressure by the rate of expansion of a pulse wave that does not contain mechanical parts has been developed. Long-term fixation of the state of the cardiovascular system for the last number of episodes of recurrence of fibrillation of the previous establishment can be applied to the abolition of the author’s mathematical model to predict possible exacerbations. c)Use the features of long-term records of a sick family member on atrial fibrillation. Interviews with patients of Vinnytsia Cardiac Dispensary were used d)A convenient software and hardware device run by a mobile application is offered. Innovative methods of collecting information about condition of the cardiovascular system, risk factors that precede the recurrence of AF are used. Such information is important for personalization of the treatment. Dynamic mathematical model of threatening forecast states is based on personal data about factors risk of AF obtained over a long period of time and current data. There is a large number of people who have atrial fibrillation but it is asymptomatic or being tolerated. This is a risk of stroke group. Even minor internal or external irritants factors can run a trigger failure mechanism rhythm. Linearly spaced pulse wave sensors give the ability to increase the accuracy of arterial measurement pressure A carbon dioxide sensor can be also used for respiratory control at night. Keywords: heart, arterial fibrillation, pulse wave, mathematical model