If the fault location of the power grid equipment only depends on the static alarm information, not records subsequent alarm information, it will leads to low positioning accuracy of fault location and multi-solution problem. So, the dynamic timing sequence of alarms are considered. Based on alarm timing sequence information, timing matrix of fault equipment state, timing matrix of alarm, timing matrix of circuit breaker, timing matrix of main protection, timing matrix of near backup protection, and timing matrix of remote backup protection, as well as method of establishing element in matrix are defined. Also, in order to solve the expected value, expected timing matrix of the circuit breaker, expected timing matrix of main protection, expected timing matrix of near backup protection and expected timing matrix of remote backup protection are defined. For forming the expected value matrix, correlation matrix of equipment-main protection, correlation matrix of equipment-near backup protection and correlation matrix of equipment-remote backup protection are defined. Based on the switch and protection information of SCADA and WAMS, the timing sequence objective function of fault location is constructed. Taking elements of the timing matrix of the device state as the quantity of state to be solved, the improved immune genetic algorithm is used to solve the timing sequence objective function. In this way, the solved device status not only includes the state of equipment failure or not, but also the time of equipment failure. Particularly, the condition that multiple devices operating at different times in power grid can clearly be identified. At the same time, the immune genetic algorithm is improved to form a memory based immune genetic algorithm, which solves the problem that the dynamic alarm information is ignored and improves the speed and accuracy of fault location. This work is supported by National Natural Science Foundation of China (No. 51774088).