Cycle slip detection and repair are essential quality control steps in recovering the integer ambiguities when loss of tracking signals in GNSS precise positioning occurs. In this contribution, we present an improvement to the previous algorithm for reliable real-time cycle slip detection and repair of the Australian Analysis Centre Software (ACS) Pre-processing and Data Editing (PDE) function. First, the traditionally used algorithm based on the quality control theory is used to detect and repair the cycle slips. Then if the cycle slips are detected but not reliably repaired, the information of subsequent epochs are used together to strengthen the model and achieve a higher cycle slip repair success-rate. With such an enhancement, the model becomes more robust to accommodate the measurement noise and the ionosphere disturbance.