Vibration-energy harvesting is an effective strategy for replacing batteries and provides a long-term power supply to microelectronic devices. Harvesting vibration energy from human motions has attracted research attention in recent years. Here, a novel low-frequency hybrid piezoelectric and electromagnetic broadband harvester is proposed. Two parallel piezoelectric cantilever beams support the harvester and capture environmental vibration energy based on the piezoelectric effect. A permanent magnet is connected by springs to the two beams, and a fixed coil surrounds the moving permanent magnet, enabling energy conversion via the electromagnetic effect and the proof mass. The parameters influencing the output power of the harvester are optimized numerically to boost the harvester’s performance. The output power of the proposed hybrid harvester is compared with that of a piezoelectric harvester and an electromagnetic harvester. The simulation results show that the output power is significantly higher for the hybrid harvester than for the piezoelectric and electromagnetic harvesters, and the bandwidth is broader owing to the double cantilevers. An experiment is conducted using a prototype of the hybrid harvester to evaluate its output power. The results show multiple resonant peaks, an extended bandwidth, and a maximum power of 6.28 mW. In contrast, the maximum harvested power of the piezoelectric harvester is only 5.15 mW at 9.6 Hz.
When a short circuit occurs in the power supply area of a distribution network with a high-permeability distributed generation, the line current will increase, the voltage will drop sharply, and the fault characteristics will be more complex. Therefore, the automatic, quick, and accurate location of fault sections is of great significance to the reliability of power supply. In order to prevent large-scale power outages in the power supply area caused by the failure of feeders and transformers, this paper proposes a novel method to locate fault sections in active distribution networks by taking into account the data of the power supply area. On the basis of the synchronization of calculated and measured time and the observability of the fault state, a limited number of intelligent terminals are reasonably arranged in the distribution network feeder and power supply area. Additionally, the fault location equation is established based on the three-phase voltage change values of the nodes before and after the fault collected by intelligent terminals, so that the fault section is determined by comparing the residuals. Finally, the proposed method is verified by the improved IEEE 37-node and IEEE 123-node three-phase distribution networks. The results show that it has high accuracy in locating fault sections in multiple fault scenarios.
Accurate online estimation of the state of charge (SOC) and state of health (SOH) of lithium-ion batteries during charging and discharging processes remains a formidable challenge in Battery Management Systems (BMS). In response to this issue, an effective online estimation method is proposed in this work. Initially, a second-order RC equivalent circuit model is established. Subsequently, to fully adapt to the time-varying characteristics of the model's internal parameters, an improved recursive least square method is integrated to realize the online identification of the model parameters. Ultimately, according to the model parameters identified online, the state equation and the observation equation under the Dual Extended Kalman Filter (DEKF) are established, and the battery capacity is substituted into the DEKF algorithm as a parameter vector to realize the online estimation of the SOC and SOH. By simulating the Federal Urban Driving Status (FUDS), the estimated error of SOC and SOH are, respectively, only 1.14% and 1.55%.