<p>It is critical for the realization of artificial intelligence that the human brain learns and remembers in unknown environments. The Lobe Component Analysis-based developmental network, which has the structure of autonomous development and the mechanism of simulating synaptic signaling, is a suitable model to study the simulation of brain intelligence. However, the existing models of developmental networks lack neural regulation mechanisms linked to different functional regions of the brain, cannot simulate the decision generation of the whole brain nervous system. They have poor adaptability and weak learning ability in variable and unknown environments. Considering these limitations, this study proposes a developmental network model that simulates the human brain’s learning and consolidation of new knowledge in unknown environments by combining the consolidation mechanism of memory and the physiological basis of the brain’s neuromodulation mechanism, which can guide the robot’s behavioral decision-making to adapt to unknown environments more quickly. The superiority of this model is verified by the robot navigation experiment.</p>