A novel exponential‐type anti‐windup method applied to the power converter in marine electrical propeller system

Abstract

This paper proposes a novel exponential anti‐windup method for a high‐proportional‐load marine electrical propeller system, allowing the system to respond to changes in working conditions that would make the control system operate in the open‐loop state and severely affect the system stability and dynamic response speed. First, the commonly used inverse calculation type and conditional integration type anti‐windup methods are discussed, and the two methods are compared and analysed. The proposed exponential anti‐integral saturation method is obtained by introducing an exponential link in the PI controller to weaken the effect of the integral link when the controller output is saturated to improve the control performance of the traditional anti‐integral saturation method when the operating conditions of the marine electrical propeller system change. To verify the accuracy of the theoretical analysis, experiments were conducted using starsim. The experimental results verify that the theoretical analysis and parameter design are correct and superior and have the characteristics of fast desaturation and small overshoot of the system. The method has good theoretical research significance and practical application prospects for improving the comprehensive performance of marine electrical propeller systems.

Comparative Investigation of DSP-Based Speed Control of PMSM Using Proportional Integral and Takagi-Sugeno Fuzzy Logic Controller

“…In the context of a PI controller, windup represents an extensively researched issue [24,25]. This problem typically arises when the input error to the controller is significant, or when the input error remains non-zero for an extended period.…”

“…Therefore, input and output gains are introduced to adjust the scaling and sensitivity of the TS-FLC (25). These gains allow for fine-tuning the influence and responsiveness of the input and output variables within the control system [26].…”

“…The scaling factor G E can be calculated (25) by considering the range of speed references from minimum to maximum [200, 200] Hz flux rotation frequency ([6000, 6000] rpm): (25) The calculation of the scaling factor (27), , for the change in speed error involves considering the motor's rated inertia , nominal torque (26), and the sampling time :…”