Research On Gait Movement And Environment Adaptation Based On CPG Network

After a long time of study, the conclusion that the rhythm movement of the animal is controlled by the central pattern generator of spinal cord has been generally recognized by biologists. Biological rhythm of movement simplifies the complex mathematical models in the traditional sports control method. The control strategy of biped robot motion is always a problem of robot motion. Therefore, this article uses the CPG model to study the bipedal robot movement. And select the Rayleigh nonlinear oscillator as the CPG controller. The nonlinear oscillator has the advantage of simple structure, stable and have inherent synchronization feature, can make the limbs coordinate.In order to verify the CPG which constructed with Rayleigh nonlinear oscillator can be applied to gait motion planning, the use of CPG model that constructed with Rayleigh nonlinear oscillator to simulate multiple neurons were studied in this paper. In this paper, relatively simple Rayleigh oscillator is used as the CPG controller to build a network model. In view of this method in the coupled neurons lack of description, established the three neuron model based on the single neuron and double neuron model to respectively simulate hip, knee and ankle joints, thus achieving the simulation of legs of each joint in the process of gait motion. Simulation results show that the CPG which constructed with Rayleigh nonlinear oscillator can well simulate physiological characteristics of many neurons. The feasibility of the Rayleigh oscillator as a CPG controller to establish multi neural network model is validated.Because of the highly nonlinear system of biped robot, the traditional control strategy will make the gait lack flexible self-organizing ability. The movement becomes very slow, stiff. Compared with the biological natural gait, there is a big difference. Nature’s creatures can use the reflection of self-excited behavior to coordinate their rhythmic movement, which can adapt to complex and changeable environment. By adding feedback information fusion and adaptive frequency oscillator learning rule, can make the above problems be good solved. When the environment changes, the CPG controller can automatically adjust the working status, thus reduce the effects of external changes, reaching a form of self-preservation and improving the organism gait motion environment adaptability.