A Workspace-Analysis-Based Genetic Algorithm for Solving Inverse Kinematics of a Multi-Fingered Anthropomorphic Hand

Abstract

Although the solution of inverse kinematics for a serial redundant manipulator has been widely researched, many algorithms still seem limited in dealing with complex geometries, including multi-finger anthropomorphic hands. In this paper, the inverse kinematic problems of multiple fingers are an aggregate problem when the target points of fingers are given. The fingers are concatenated to the same wrist and the objective is to find a solution for the wrist and two fingers simultaneously. To achieve this goal, a modified immigration genetic algorithm based on workspace analysis is developed and validated. To reduce unnecessary computation of the immigration genetic algorithm, which arises from an inappropriate inverse kinematic request, a database of the two fingers’ workspace is generated using the Monte Carlo method to examine the feasibility of inverse kinematic request. Furthermore, the estimation algorithm provides an optimal set of wrist angles for the immigration genetic algorithm to complete the remaining computation. The results reveal that the algorithm can be terminated immediately even when the inverse kinematic request is out of the workspace. In addition, a distribution of population in each generation illustrates that the optimized wrist angles provide a better initial condition, which significantly improves the convergence of the immigration genetic algorithm.

Current trend in control of artificial intelligence for health robotic manipulator

“…Improved genetic algorithm radial basis function neural network model. The optimal trajectory is solved quickly using the desired output and a genetic algorithm that enhances the radial function of the neural base network [98]- [100]. This method can greatly increase the calculation speed and guarantee real-time performance while ensuring accuracy.…”

A Boundary Scan Test Vectors Optimization Method Based on Improved GA-AO* Approach Considering Fault Probability Model

“…While keeping the testing efficiency unchanged, the above methods generate MTV by using intelligent iterative algorithm and reduce the misjudgment rate and confusion rate of MTV to some extent. However, the improper design of particle iteration and the difficulty to determine the search direction of these intelligent algorithms leads to the fast convergence in the optimization of misjudgment rate and confusion rate, the result is always trapped in the local optimal solution [20][21][22][23][24][25]. With the increase in the network size of the circuit under test, the local optimal solution of misjudgment rate and confusion rate will lead to a serious decline in the fault detection accuracy of the boundary scan [26][27][28].…”

“…At the same time, this paper sets the threshold number of searches to c s . When cumulative number of searches reach c s , this paper adjusts θ according to Equation (22).…”

“…If the ratio is equal to or less than θ, the algorithm ends, and the MTV n with least P MTV is the final solution. Otherwise, if the cumulative number of searches reaches the c s , this algorithm adjusts the θ by Equation (22).…”

New Trends in the Control of Robots and Mechatronic Systems

“…In recent years, research into the control of robotic and mechatronic systems has led to a wide variety of advanced paradigms and techniques, which have been extensively analysed and discussed in the scientific literature. Some examples of relevant approaches and methods on which researchers are focussing their efforts are fuzzy control [1,2], neural networks [3,4], sliding mode control [5], fractional-order and distributed-order control [6][7][8], reinforcement learning [9,10], genetic algorithms and evolutionary computation [11,12]. Frequently, on the basis of the specific application requirements, these techniques are used in combination, increasing the level of complexity of the control strategy [13].…”