The number of people using the web objects has enlarged quickly. This situation effects the traffic network. At present, the web cache memory is used to solve the traffic network problem. It can keep web objects from original servers and clients can download web objects from the web cache memory. However, the web cache memory has a small memory therefore it cannot record every web object. Many techniques are used to solve this problem but these techniques give the average hit rate of not over 40 percent. Therefore, this research investigates the optimization with mathematical and statistical methods for increasing the hit rate with web cache memory which are as follows: estimated value, interpolation, cubic spline, to find area under the curve function and first order condition. This research investigates to create the Repairable Least Recently Used (RLRU) algorithm to recommend for the web cache memory management. This algorithm is tested with the datasets obtained from a university in Thailand. Furthermore, the RLRU algorithm has compared the performance of the hit rate with the LRU algorithm. The experimental results of this research can apprize that the RLRU algorithm gives the maximum hit rate at 72.56 percent while the LRU algorithm offers the maximum hit rate at 53.34 percent. However, the average hit ratio of the RLRU algorithm is at 53.80 percent while the average hit ratio of the LRU algorithm amounts to about 12.99 percent.