Development of a robust intelligent controller for a semi-active car suspension system

In pursuite of comfortable driving in unpaved and off-roads, intelligent methods are used to improve the suspension systems in the vehicles. Semi-active suspension systems outperformed passive and active suspension systems because it contains an intelligent actuator that can give the appropriate force to dissipate unwanted vibration using intelligent and real-time controllers. This research examines Magneto-rheological (MR) fluid damper with a Fuzzy-PID controller, one of the most extensively intelligent semi-active suspension system's actuators researched. However, the Fuzzy logic algorithm used in the Fuzzy-PID controller cannot be wholly considered as a real-time controller; since it is fuzzy rules are designed offline and according to a previous knowledge base, which may not cope with the instant, unexpected vibrations that may occur. Commonly, the Fuzzy rules are optimized using offline optimization methods such as Differential Evolutionary (DE), Particle Swarms Optimization (PSO), or Artificial Neural Network (ANN) algorithms. In this research, Differential Evolution (DE) algorithm is modified to enhance the Fuzzy logic output gains to increase the performance of PID portion of the Fuzzy-PID controller. To ensure stability and robustness of the developed system, an active force controller (AFC) was added and tested to validate the final AFC-Fuzzy-DE-PID controller. The developed AFC-Fuzzy-DE-PID model was tested in two manners, first by simulation using MATLAB Simulink with sinusoidal and random disturbances. Then the model was tested experimentally in a quarter car test rig using different disturbances by means of a pneumatic actuator as an excitation. The test rig was developed at the control lab, Faculty Of Engineering in UPM. Results of the simulation tests for the developed controller showed that it has improved the vehicle's ride comfort by 23% - 62% better than the Fuzzy-DE-PID controller and both the Fuzzy-PID and the passive system, respectively, in sinusoidal disturbance condition. While in the random disturbance, the AFC-Fuzzy-DE-PID improved the vehicle's ride comfort by 48%, 83%, and 27% better than the Fuzzy-DE-PID, Fuzzy-PID, and the passive system, respectively. In the experimental tests on sinusoidal disturbance, the AFC-Fuzzy-DE-PID improved the ride comfort by range of 0.4% - 2% better than the Fuzzy-DE-PID, range of 6%-14% better than the Fuzzy-PID, and range of 30%-51% better than the passive system. While on the random disturbance of the experimental test, the ride comfort improved 1%, 3%, and 4% better than the Fuzzy-DE-PID, Fuzzy- PID, and the passive system, respectively. By using this developed controller in any other real-time application, it will improve the performance to the highest levels without the need for a previous knowledge base for designing a real-time Fuzzy-PID controller.

Text FK 2022 100 IR.pdf Download (1MB)

Actions (login required)

View Item