Data-based PID control of flexible joint robot using adaptive safe experimentation dynamics algorithm

Abstract

This paper proposes the data-based PID controller of flexible joint robot based on adaptive safe experimentation dynamics (ASED) algorithm. The ASED algorithm is an enhanced version of SED algorithm where the updated tuning variable is modified to adapt to the change of the objective function. By adopting the adaptive term to the updated equation of SED, it is expected that the convergence accuracy can be further improved. The effectiveness of the ASED algorithm is verified to tune the PID controller of flexible joint robot. In this flexible joint control problem, two PID controllers are utilized to control both rotary angle tracking and vibration of flexible joint robot. The performance of the proposed data-based PID controller is assessed in terms of trajectory tracking of angular motion, vibration reduction and statistical analysis of the pre-defined control objective function. The simulation results showed that the data-based PID controller based on ASED is able to produce better control accuracy than the conventional SED based method.

PID Tuning Method Using Chaotic Safe Experimentation Dynamics Algorithm for Elastic Joint Manipulator

“…The statistical performance of CSEDA was evaluated in this study using 25 independent trials. Besides, the performance was compared among the proposed CSEDA and the original SEDA, ASEDA [25], and LFSEDA [26].…”

“…Thus far, trajectory-based self-tuning methods that are commonly used to determine the optimal controllers include simulated annealing (SA) [19], random search [20], simultaneous perturbation stochastic approximation (SPSA) [21], and safe experimentation dynamics algorithm (SEDA) [22][23][24]. Among these trajectorybased self-tuning tools, SEDA is the most significant one due to its memory-based structure, simplicity, and fewer coefficient values [22][23][24], and it has often been used to tune the PID controller of an elastic joint manipulator [25,26].…”

“…This paper proposed to use adaptive SEDA (ASEDA) [25] and levy-flight-based SEDA (LFSEDA) [26] for improving the convergence accuracy of the original SEDA. However, less exploration was produced in the initial searching stage because their updated equations still used the fixed value of step size gain or interval size.…”

Optimal integral sliding mode controller controller design for 2-RLFJ manipulator based on hybrid optimization algorithm

“…This has led to much research into developing high-performance control approaches using state-of-the-art control theories [1]. For instance, proportional integral derivative (PID) controller [2]- [4], sliding mode control [5], fractional-order sliding mode controller [6], adaptive sliding mode control [7], fuzzy sliding mode control [8] have been dedicated to the study of flexible-joint robots. An integral sliding mode controller (ISMC) tracks a flexible joint manipulator driven by a direct current (DC) motor.…”

Switched time delay control based on neural network for fault detection and compensation in robot

“…There are different learning approaches which differ from each other by the way of adjusting the weights and their structure depends on the architecture of the neural network and the task to be performed. Besides, neural networks have been searched and carried out in real systems [8,9]; there are many ANN applications in data analysis, identification, and model control [10]. Amid various types of ANN, a multi-layer perceptron (MLP) is quite popular and used extensively in research.…”