Speed tracking of Brushless DC motor based on deep reinforcement learning and PID
Puwei Lu, Wenkai Huang, Junlong Xiao
Abstract
In order to improve the speed tracking accuracy of Brushless DC motor, this paper proposes a PID control method based on deep deterministic policy gradient algorithm(DDPG) compensation. The running state of BLDCM is monitored by DDPG, and the proportional, integral and differential links of PID are compensated to make PID controller have learning ability. At the same time, the Gauss function and the maximum critical speed interval are combined to define the reward function to enhance the versatility of the control strategy. Simulation results show that the proposed strategy has better speed tracking accuracy than PID controller.
Topics & Concepts
PID controllerControl theory (sociology)DC motorReinforcement learningTracking (education)Computer scienceCompensation (psychology)Controller (irrigation)Control engineeringTorqueArtificial intelligenceEngineeringControl (management)Temperature controlPhysicsPsychologyThermodynamicsElectrical engineeringPedagogyAgronomyBiologyPsychoanalysisSensorless Control of Electric MotorsElevator Systems and ControlAdaptive Dynamic Programming Control