Dynamic Stdp Synapse System Modeling Research And Validation

STDP (Spike Timing Dependent Plasticity) mechanism of synapses is considered to be one of the most important mechanisms of cerebral neural network, and the synaptic mechanisms of STDP should be taken into account in modeling of large-scale cerebral neural network. The most effective method of emulated STDP synapse is to establishing the kinetic model of synaptic ion channels.However, most of the existing modeling of STDP synaptic mechanism have some certain deficiencies which could be divided into two aspects. On the one hand, they either can’t well explained the biomolecular reactions at a cellular level which is responsible for STDP, or acquired asymmetric STDP responses. On the other hand, there is no perfect proof scheme which could be used to validate the modeling. For the first question, a new modeling of STDP synaptic system was proposed; with regard to the second one, we proposed an ionic channel kinetic model of glutamate signal and back propagation dendrite signal, in order to model the pre-synaptic glutamate signal and post-synaptic action potential accurately, and then used those two signals to make the verification of the STDP synaptic system. Therefore, this thesis includes the following:Firstly, the thesis briefly introduces the main components of the cerebral neural network and the principle of STDP synaptic mechanisms, analyses the classic Hodgkin-Huxley neuron model and does the simulation and testing for the model.Secondly, the thesis modeling the two input signals of the STDP synapse-glutamate signals and dendritic signal, analyses the biomolecular reactions at a cellular level which is responsible for these two signals, and then structures ionic channel kinetic models of these two signals respectively. We do the simulation of those kinetic models and validate the corresponding results later.Thirdly, the thesis studies the internal structure of the STDP synapse. We proposes NMDA receptors and control model of Ca ionic signal which constitute the internal structure of the STDP synapses, analyses the impact of these two structures on the connection strength of the STDP synapse. Finally, the thesis conducts comprehensive simulation and verification of the system model of STDP synapse. The simulation results show that the new modeling not only reflects the actual physiological mechanism of STDP synapses, but also asymmetric STDP responses are obtained. Likewise, the verification results of proof scheme not only get the actual physiological glutamate signal and dendritic signal, but also obtains asymmetric STDP responses from the verification of STDP synaptic model.