Pulse neural network training method based on membrane potential self-increasing mechanism

The invention discloses a pulse neural network training method based on a membrane potential self-increasing mechanism, relates to a pulse neural network training method, in particular to a time coding training method based on the membrane potential self-increasing mechanism, and belongs to the field of artificial intelligence. When a pulse neural network is trained, due to the influence of signal sparsity, only a few neurons are activated, and errors of a network output layer cannot be effectively propagated to each hidden layer of the network and cannot participate in parameter updating. In order to solve the problem, according to the invention, when a time coding method is adopted to train a pulse neural network, a self-increasing item changing along with time is added to a pulse neuron cell membrane potential dynamical model. The self-increasing item enables all the pulse neurons to be activated in limited time, so that parameters of all the neurons can be updated in the back propagation process, and the