The problem of iterative learning control for a class of irregular distributed parameter systems is considered. The considered distributed parameter systems are composed of parabolic partial differential equations. According to the characteristics of the irregular systems, iterative learning control laws are proposed for such irregular distributed parameter systems by using the D-type learning scheme. Convergence of the output tracking errors in ??2 norm along the iteration axis is proved based on the contraction mapping method. A simulation example shows the feasibility and effectiveness of the obtained conclusion.