Time sensitive networking(TSN) has the advantage of deterministic low latency

making it highly compatible with the strong real-time and highly reliable communication requirements of substation automation. To address the problem of filtering abnormal traffic in substation networks

we propose a method for filtering abnormal traffic in time-sensitive networks of substations based on credit values. At the input of the switch

we propose an improved per stream filtering and policing(PSFP) structure

which gradually filters out abnormal traffic in the input stream

including frames exceeding the length threshold

frames not arriving at the specified time

and frames exceeding the preset bandwidth limit

through the use of stream filters

flow gates

and credit-based flow limiters. On this basis

we propose a credit-based flow limiting algorithm

which configures parameters such as idle slope

send slope

and maximum credit based on the bandwidth and maximum burst size of the business flow. This algorithm filters out abnormal network traffic using credit value. Finally

we build a multi-bay substation communication network model and conduct simulation experiments. The experimental results show that the proposed method restores the heavily occupied bandwidth resources of the switch

reduces the delay and jitter of each business flow

and ensures the reliable transmission of each flow in substation during network storm.