Study On Local Field Potentials Of Cerebellum And Midbrain Of Carp For Biological Control

Bio-robotics are an emerging frontier science and technology field in today’s world Which multiple disciplines are cross-fertilized.Bio-robots can replace humans to accomplish complex and dangerous tasks in the future,so bio-robots have important scientific research significance and broad application prospects.Aquatic bio-robotics are an important branch in the field of bio-robotics,which has unique advantages and significant features in practical applications.In order to investigate how to effectively control aquatic bio-robots for long periods of time by studying local field potentials in the cerebellum and midbrain,this thesis is an experimental study using carp as the object of this research:Since carp maintain a resting state or slow swimming state for a longer period of time to reduce the metabolic level of the body to reduce energy consumption,in order to capture the EEG signals in the steering state,this thesis designs a photoelectric stimulation piggyback device with a rotatable light source for the control of the locomotor behavior of carp robots,and uses a homemade photoelectric stimulation combination device and a biosignal acquisition and analysis system to realize the acquisition of EEG signals in different brain regions at rest.The local field potential signals were recorded in the cerebellar motor area of the carp brain and the midbrain optic cap using a combination of a homemade photoelectric stimulation device and a biosignal acquisition and analysis system to acquire the carp EEG signals in different brain regions at rest and in the same motor behavior under different brain regions with light stimulation.The acquired local field potential signals were affected by certain background noise.Therefore,in this thesis,signal processing was performed on the EEG signals of carp underwater in the resting state and in the motion state induced by light stimulation,using digital filtering,wavelet denoising and wavelet packet decomposition to remove baseline drift to obtain signals with more useful signal content and less noise signal content.The processed EEG signals were then analyzed using time-frequency analysis.In this thesis,the local field potential signal was decomposed into five sub-band frequency bands,and wavelet packet energy was calculated separately for each band to extract features.The results show that the components of , and frequency bands increase during the transition from resting state to light stimulation-induced locomotion state in carp.This study speculates that these three frequency bands may contain more motor information.Furthermore,comparing the wavelet packet energies of the cerebellum and midbrain,it was found that in the and frequency bands,the energy of the cerebellum was somewhat smaller than that of the midbrain,while in the and frequency bands,the energy of the cerebellum was again smaller than that of the midbrain.The changes in the and frequency bands of the cerebellum and midbrain were greatest in the motor state,where the changes in were particularly pronounced by light stimulation.In this thesis,the EEG signals of the carp cerebellum and midbrain under the same locomotor behavior were studied using a three-fold technical approach in the time domain,frequency domain and time-frequency domain.The results showed that the EEG signals of the carp cerebellum and midbrain are correlated and differential.In the resting state,the correlation between the carp cerebellum and midbrain was higher,and the relative power of each frequency band was similar,with the cerebellum containing somewhat higher information than the midbrain.In the light stimulation-induced locomotion state,the correlation between the carp cerebellum and midbrain decreased,and the relative power of each frequency band appeared to be different.The changes in both cerebellar and midbrain and frequency bands are larger,and the changes in the midbrain frequency band will be a bit stronger,and the information contained in both brain regions has increased,and both the carp cerebellum and midbrain are involved to different degrees throughout the locomotion process.To investigate the changes in the locomotor behavior of the carp robot when the stimulation frequency was changed,the present study divided the carp underwater stimulation experiment into a cerebellar stimulation group,a midbrain stimulation group,and a combined cerebellar and midbrain stimulation group.The experimental results of the single stimulation group showed that applying electrical stimulation at a frequency of about35 Hz to the cerebellar motor area resulted in the highest control success rate of 65%.In contrast,electrical stimulation of the midbrain motor area at about 40 Hz was required to achieve the highest control success rate,which could reach 70%.The results of the combined stimulation group showed that the combined stimulation was beneficial in increasing the control success rate of the carp robot,but the control accuracy was not as stable as that of the single stimulation,and more intense movements of the carp occurred.This study provides a new strategy for precise and effective control of carp robots.