近年来，类脑计算引起国内外研究者的高度重视。为抢抓类脑计算发展的重大战略机遇，构筑我国人工智能发展的先发优势，“中国脑计划”已作为重大科研项目列入国家“十三五”规划。本课题在神经网络的最新研究成果上，结合生物神经元的微柱结构等认知机理，研究神经元优化组合策略与模块特殊功能间的有效匹配机制，借鉴不同脑区模块间的协同处理模式，结合记忆、推理和注意等认知原理，研究不同神经元模块之间信息有效传递方式，构建跨尺度信息处理机制，使得不同神经元模块间传递的信息能够有效地融合，以实现复杂的不同认知功能。具体而言，是研究适用于模块功能特异化的Spike编码方案、新型具有模块功能特异化性质的Spiking神经网络模型以及适合新型Spiking神经网络模型的高效学习算法。以突破现有脑科学中神经网络研究的瓶颈，从更深层次上探索人脑智能的本质，大幅提升网络的认知功能。

In recent years, brain-inspired computing has attracted wide domestic and overseas attention in the academic community. In order to grasp the great strategic opportunities for the development of the brain-Inspired computing, and to build the first-mover advantage of China's artificial intelligence development, China Brain Blan” has been included in the national 13th five-year plan as a major scientific research project. On the basis of the latest research findings, this project intends to carry out the optimal combination strategy of neurons and the effective matching mechanism between modules’ special functions by combining the cognitive mechanisms such as micro-column structure of biological neuron. Learning from the cooperative processing mode between different brain regions and combining with the cognitive principles such as memory, reasoning and attention, this project studies the effective information transmission way between different neuron modules, and constructs the cross-scale information processing mechanism, so that the information transmitted between different neuron modules can be effectively fused to achieve complex and different cognitive functions. More precisely, spike coding schemes suitable for module function specialization, new model of spiking neural network and efficient learning algorithm suitable for this new model are studied, in order to break through the bottleneck of existing neural network research in brain science, explore the essence of the human brain intelligence from the deeper level, promote the cognitive function of the network greatly.