The article proposes a chaos neural network model applied to autoassociation memory. The presented artificial neuron model is properly characterized in terms of a time-dependent skew-tent periodic activation function to involve a chaotic dynamics as well as the energy steepest descent strategy. It is elucidated that the present neural network has a remarkable performance of dynamic memory retrievals beyond that of conventional models with nonmonotonous activation function as well as a monotonous activation function (e.g., sigmoidal). This advantage is found to be attributed to the property of analogue periodic mapping accompanied by chaotic behaviour of the neurons. It is concluded that the presented analogue neuron model with periodicity control has an apparently large memory capacity in comparison with previously proposed association models.