Search

Multi-agent collaborative perception (MCP) has recently attracted much attention. It includes three key processes: communication for sharing, collaboration for integration, and reconstruction for different downstream tasks. Existing methods pursue designing the collaboration process alone, ignoring their intrinsic interactions and resulting in suboptimal performance. In contrast, we aim to propose a Unified Collaborative perception framework named UMC, optimizing the communication, collaboration, and reconstruction processes with the Multi-resolution technique. The communication introduces a novel trainable multi-resolution and selective-region (MRSR) mechanism, achieving higher quality and lower bandwidth. Then, a graph-based collaboration is proposed, conducting on each resolution to adapt the MRSR. Finally, the reconstruction integrates the multi-resolution collaborative features for downstream tasks. Since the general metric can not reflect the performance enhancement brought by MCP systematically, we introduce a brand-new evaluation metric that evaluates the MCP from different perspectives. To verify our algorithm, we conducted experiments on the V2X-Sim and OPV2V datasets. Our quantitative and qualitative experiments prove that the proposed UMC greatly outperforms the state-of-the-art collaborative perception approaches.

Vehicle-to-grid reserve means that an electric vehicle could serve as a reserve power supplier to meet the reserve demand of a power company through providing the electricity of batteries back to the grid. However, this new reserve would change the current reserve dispatch model because of the heterogeneity between the vehicle-to-grid reserve and the generation-side reserve in some properties. A new operating paradigm taking the vehicle-to-grid reserve into account is proposed here after summarizing the market properties that the vehicle-to-grid reserve brings. The detailed operational steps are formulated containing reserve demand forecasting, selecting criterion setting, and the corresponding scheduling strategy. Finally, the effectiveness and validity of the proposed paradigm is illustrated by a numerical example

In this paper, the artificial contamination test of composite insulator(FXBW-110/100) was carried out in the artificial climate chamber. The leakage current signal of composite insulator with different contamination and hydrophobic state was collected. Through the analysis of the leakage current of the time and frequency domain, its characteristic parameter characterization method of leakage current based on the pulse amplitude(Ih), energy ratio(K) and energy(E) of leakage current is proposed. The results show that under the same voltage, the leakage current pulse amplitude increases with the increase of salt deposited density (SDD) of composite insulators. With the increasing of SDD, there are more frequent the surface discharge occurred and more spike pulses in the leakage current waveform. In the same pollution state, with the decline of the hydrophobicity(HC) grade of composite insulators, the amplitude of the leakage current increases gradually and the frequency of the pulse increases gradually. Through the Fast Fourier Transformation(FFT) analysis of the leakage current, it is found that the highest leakage current is the basic wave, the three and the five harmonic. With the increasing of SDD and hydrophobicity, all of the leakage current harmonics were increased to a certain extent, however, the increasing amplitude of three times and five times harmonic were bigger than that of fundamental wave. The three characteristic parameters of leakage current, i.e. pulse amplitude (7h), energy ratio(K) and energy (E) all increase with the increase of SDD, and the decrease of HC.

A de-noising method based on the combination of wavelet and Ensemble Empirical Mode Decomposition (WEEMD) theory is proposed for de-noising of leakage current signal. Results show that the proposed method is effective to extract leakage current signals. The noise in the leakage current get by a non-inductance resistance with 1 ohm is very large, and the noise signal almost drowns the leakage current signal. After de-noising, the denoising effect of the leakage current signal is not ideal, and the characteristics of the leakage current itself are lost. Because of the higher signal-to-noise ratio (SNR) of the leakage current which measured by 5 ohm and 50 ohm non-inductance resistance, the de-noised effect is much better. Therefore, the hardware measures are taken to reduce the noise contained in the leakage current signal, and then the clean leakage current signal can be obtained by de-noising method proposed in this paper.