Relevance. Currently, the majority of oil production is carried out using centrifugal electric pumps, and the share of installations operating in periodic modes with cycles of accumulation and pumping of liquid is increasing every year. At the same time, the latest studies of the reliability of their components show that the mean time between failures of submersible equipment is reduced to 2.5 times compared to equipment in continuous operation modes. The decrease in reliability is associated with the dynamics of transient processes in automated lower-level control systems, which is currently not taken into account when developing management decisions to optimize well operation, and the entire lower-level subsystem, as a rule, is represented by linear or nonlinear static links. This approach does not allow controlling the dynamics of processes within the subsystem itself in the conditions of wells being transferred to intermittent operating modes. This determines the relevance of developing dynamic models of the submersible pump electric drive subsystem in order to optimize transient processes according to the criterion of the time between failures of submersible equipment elements. Aim. Improving models of lower-level automated control systems for a submersible electric motor of a centrifugal electric pump installation for oil production under conditions of switching a well to periodic operation mode in order to improve controllability. Methods. Methods of system analysis and structural-parametric synthesis of automated control systems of technological processes, methods of designing vector control systems for electric drives, methods of identifying state variables based on observers using explicit mathematical models, methods of parametric identification of customizable models of non-stationary dynamic systems. Results. The authors have proposed the approach to constructing an operational control level for centrifugal electric pump installations, including the introduction of closed sensorless electric drive systems using state observers based on explicit adjustable mathematical models of the non-stationary dynamic system "long cable – submersible adjustable electric machine" and methods of their dynamic identification, creating prerequisites for further formulation and solution of problems of optimizing dynamic processes within the lower-level subsystem under conditions of transferring the well to intermittent operating modes. A transition to dynamic models of automated lower-level control systems based on reduced mathematical models of the electrical subsystem of centrifugal electric pump installations is proposed for increasing controllability and taking into account the physical limitations of available measuring instruments.