Skip to main content

Table 1 Pros and cons of existing protection schemes

From: Current trajectory image-based protection algorithm for transmission lines connected to MMC-HVDC stations using CA-CNN

Protection scheme

Pros (merits)

Cons (demerits)

Frequency-Domain based scheme [7,8,9, 14]

The fault behaviors of MMC are considered. Thus, they can operate well in the presence of MMC-HVDC stations

Inherent demerits of frequency-domain based scheme include the requirement of discrete Fourier transform, the negative effect of decaying DC component on accuracy of phasor estimation

Some of them have no natural phase selection capability [7,8,9]

Time-Domain based scheme [10,11,12,13, 15]

Discrete Fourier transform is removed. They exhibit good operating speed and low complexity

Some of them are not applicable to MMC-HVDC stations. Furthermore, they do not have excellent tolerance against some nonideal conditions, and therefore may operate incorrectly during an external fault in the presence of CT saturation, and outliers [10,11,12,13]

Operating performance depends heavily on the reasonable selection of protection threshold [15]

Deep learning- based scheme [16,17,18,19,20,21]

Existing solutions that combine deep learning with relay protection focus on two aspects: fault classification [16,17,18] and fault location [19,20,21]. Deep learning has strong nonlinear fitting ability, feature expression ability

The fault behavior of the MMC-based converter station is not considered, and the proposed method may have adaptability problems for lines connected to MMC-HVDC. Some of them can only detect ground faults without considering other fault types [19,20,21]