Algorithm of fuzzy sliding mode iterative learning control was proposed for repetitive nonlinear rolling mill temperature control system with uncertainties. The algorithm was an iterative learning control algorithm, which took the function of a sliding mode control as the input of fuzzy control and took the fuzzy control output as the control increment of iterative learning control. To improve the convergence rate of ILC, a new ILC based on a sliding mode fuzzy control was presented which combined a sliding mode control with ILC. The fuzzy control in the algorithm can smooth the control signal and lessen the common dithering of sliding mode control. The simulation results indicate that the new algorithm is effective, which can achieve higher speediness than simplex feedback control.

In response to the spam problem, the present mail transfer protocol-SMTP was analyzed. As the commands are transmitted by plaintext in the SMTP, it is easy to bring forth some hidden trouble of safety. So the SMTP was extended in its type of data and credibility. Based on the result, a new model was set up and its realization was tested. The result of test shows that the case is effective.

The technology of BP neural network was applied to the predictive control algorithm which was studied. The BP neural network was used to implement identity and multistep prediction. A new type of nonlinear chaotic map was introduced into the learning algorithm of neural network parameters in order to realizing the weight regulation. The weight of the network was modified to create a reasonable multistep predictive model. Control law was modified by the deviation between the multistep predictive outputs sequence of the network and the set point sequence of the system. Simulation results show that the proposed method is effective.