this paper concerns the asymptotic synchronization of delayed reaction-diffusion neural networks (rdnns) with unknown nonidentical time-varying coupling strengths, where the time-varying coupling strengths are consist of continuous time-varying periodic parameters and time-invariant nonnegative parameters. by utilizing a novel adaptive approach, the differential-difference type adaptive laws of coupling strengths and adaptive controller are designed such that the nonidentical rdnns are asymptotic synchronization. the sufficient conditions dependent on the reaction-diffusion terms are derived by constructing a novel lyapunov-krasovskii-like composite energy functional (cef) and using barbalat's lemma. finally, a simulation example is provided to illustrate the effectiveness of the developed approach. (c) 2016 elsevier b.v. all rights reserved.