SVM-based Gait Control Method For Hydraulic Quadruped Robots

Abstract

With the development of technology, there are more and more researches on quadruped robots in the field of artificial intelligence, but there is a lack of research on gait control methods of quadruped robots. To solve this problem, a GA-SVM algorithm model is proposed by combining Support vector machine (SVM) and Genetic Algorithm (GA). This model optimizes the important parameters of SVM through GA, and improves the performance of SVM. Then compare it with neural network (NN) model and traditional SVM model to verify its performance. The experimental results show that the value of GA-SVM model in the training set is 0.9215, and its performance is better than that of traditional NN model and traditional SVM model. In the simulation test of gait control system, GA-SVM model is 9.391, which is better than traditional NN model and traditional SVM model. The results show that the performance of GA-SVM model obtained by the combination of GA and SVM has been greatly improved compared with the traditional SVM, which can provide a new idea and method for the gait control of quadruped robot.

Keywords

• ga
• gait control
• quadruped robot
• svm
• artificial intelligence