Non-intrusive load monitoring (NILM) is the practice of estimating power consumption of a single household appliance using data from a total power meter of the user’s house. The transformer model has emerged as a popular method for handling NILM problems. However, with the increase in data from electricity meters, there is a need for research focusing on the accuracy and computational complexity of the transformer model. To address this, this paper proposes a sequence-to-sequence load decomposition structure named GTCN, which combines the gate-transformer and convolutional neural networks (CNNs). GTCN introduces a gating mechanism to reduce the number of parameters for training the model while maintaining performance. The introduction of CNNs can effectively capture local features that the gate-transformer may not be able to capture, thereby improving the accuracy of power estimation of individual household appliances. The results of the experiments, based on the UK-DALE dataset, illustrate that GTCN not only demonstrates excellent decomposition performance but also reduces the model parameters compared to conventional transformers. Moreover, the proposed GTCN structure, despite maintaining the same number of model parameters as the traditional transformer architecture after incorporating CNNs, outperforms the conventional transformer model, as well as current seq2seq and R-LSTM technologies, and achieves enhanced prediction accuracy and improved generalization capability.

This study addresses the design of a new type of permanent magnet contactor, which uses forced breaking to overcome failure conditions. A multi‐objective optimisation method based on the genetic algorithm is used to design the contactor considering the effects of average closing time, average breaking time, impact energy, Joule heat and contactor volume. Four prototypes have been fabricated according to the optimised parameters. A control circuit is developed to increase breaking reliability. The circuit is designed as two separate modules: a normal working module and a protection module. The former is used during the normal contactor operation and the latter is used for forced breaking protection. Breaking failures are detected by comparing the time delay between the breaking control voltage and the auxiliary contact voltage. A durability experiment has been conducted to verify the protection function of the proposed PM contactor. Experimental results show that the normal working module fails to break the contactor in the last testing cycle. However, under the actuation of the protection mechanism, the contactor is forced to break and the power circuit can be successfully switched off.

Circuit breakers (CBs) have played important roles in the power distribution network. According to IEC standards, the protection characteristics need to be investigated in the manufacturing process to ensure high reliability. It is required from IEC 60947 that the instantaneous release shall be verified with a test current at least 10 times of the rated current. For large-capacity CB testing, converters with large output current, high accuracy and fast response are needed. A large current indirect converter including four converting units and a matrix transformer is proposed. The converting units are synchronized using a CPLD based logic control unit. To increase the current accuracy, an adaptive feedforward control strategy is developed to compensate the influences of CB-dependent resistances and inductances. The equivalent load impedance is adaptively identified from a robust identifier, which is then used to construct the feedforward compensator. Simulation and experimental results show that different circuit breakers bring in CB-dependent contact resistances and inductances in the testing circuit. The impedance change can be successfully identified by the robust identifier with limited errors. The proposed converter is capable of generating a wide range of current from 100 A to 40 kA with the error no more than 2.2%.