The research is considered relevant as there is a need for higher reliability and lower fault rate of asynchronous electric drives caused by the failure of the stator winding. The research aims to develop a method to account for the effect of winding temperature changes under load fluctuations on the thermal resource of winding insulation. The investigations were carried out using a 5-node thermodynamic model of a TEFC asynchronous motor operating in the random loading mode. The motor series was clustered with the subsequent creation of a generalized thermodynamic model for each cluster. The parameters of the generalized model were determined with the genetic algorithms method. The simulations based on generalized thermodynamic models allowed generating the expressions for approximate accounting of the impact that the temperature fluctuations have on the acceleration of the insulation thermal resource expenditure. The paper suggests a simple method of accounting of this influence on the stage of selection and verification of the motor during the electric drive design.