Wang Dong

Bio: Wang Dong is an academic researcher. The author has contributed to research in topics: Fault (power engineering) & Photovoltaic system. The author has an hindex of 1, co-authored 6 publications receiving 1 citations.

Photovoltaic Power System Fault Warning based on State Assessment

Abstract: The stability of photovoltaic power system is very important. A small fault in one component can cause a large damage on the whole power system. This paper aims to propose a method to evaluate the stability of photovoltaic power system so that we could take measures in advance to avoid greater losses. This paper studies how we assess the status of a power grid. To assess the status of power grid, we select some fault which is often occur in photovoltaic power system. Then we build a fault tree based on the photovoltaic power plant network. Finally, we substitute the probability of various faults into the fault tree for analysis and prediction. According to probability of failure at this moment and coefficient, we can calculate how the system running at this moment. In this paper, the early fault warning of photovoltaic power system: firstly use machine learning methods to pre-process the original data, eliminate redundancy and noise, then run regression analysis on pre-processed data and get the failure rate of a certain time. Take the example of a photovoltaic power system in China, through learning and verification of fault data, our method can effectively realize the early fault warning of photovoltaic power system.

An Intelligent Fault Diagnosis Technology for Distributed Photovoltaic System Based on Ensemble Voting Fusion Model

Abstract: With the gradual consumption of fossil energy, the development and use of renewable energy become more and more important. Distributed photovoltaic power stations have become an important means to efficiently use solar energy for power generation. Under the influence of complex environmental factors, with the continuous expansion of photovoltaic power stations, photovoltaic system accidents occur frequently, and problems in operation and maintenance gradually emerge. In order to improve the informatization level of photovoltaic system and reduce operation and maintenance costs, this paper designs an ensemble model based on cross-validation, which integrates five classical classification algorithms. Through analyzing and processing multidimensional data of photovoltaic system, it realizes the function of fault diagnosis and fault classification of photovoltaic system. By comparing the performance of single model and ensemble model in photovoltaic fault diagnosis, it is found that the ensemble model has better diagnostic performance and can be used for photovoltaic system operation and maintenance and fault diagnosis.

Distributed photovoltaic power station fault monitoring method and device based on edge computing

Abstract: The application provides a distributed photovoltaic power station fault monitoring method and device based on edge computing, which are applied to an edge computing device. The method comprises the steps of: obtaining meteorological and environmental parameters of an area where a target distributed photovoltaic power station is located, cloud block image data and field equipment image data above the target distributed photovoltaic power station; predicting the power generation power of the target distributed photovoltaic power station in a set time period by using the meteorological and environmental parameters, the cloud block image data above and the field equipment image data to serve as predicted power generation power; acquiring actual power generation power of the target distributedphotovoltaic power station within the set time period; and monitoring whether the target distributed photovoltaic power station has abnormity or faults or not by comparing the predicted power generation power with the actual power generation power. In the application, the abnormity or fault monitoring of the target distributed photovoltaic power station can be achieved through the mode, and the monitoring reliability is improved.

Abnormal data detection method and device for photovoltaic power station and electronic equipment

Abstract: The invention provides an abnormal data detection method and device for a photovoltaic power station and electronic equipment. The method comprises the steps of after obtaining photovoltaic residual data corresponding to a power station, determining a clustering center point of the photovoltaic residual data based on the photovoltaic residual data; clustering the photovoltaic residual error data on the basis of the clustering center point to obtain a clustering result; determining the photovoltaic residual data which does not belong to any clustering cluster as abnormal data. According to themethod and the device, the abnormal data can be determined, so that when the abnormal data occurs, the fault handling is carried out in time, and the stability and the safety of an electric power system are ensured. Furthermore, the clustering center point is determined through the density value dimension data and the distance value dimension data, so that the determined clustering center point ismore accurate, the clustering result obtained by using the clustering center point is more accurate, and the determined abnormal data is more accurate.

Photovoltaic system power generation state monitoring method and device

Abstract: The invention provides a photovoltaic system power generation state monitoring method and device, and the method comprises the steps: firstly determining a benchmarking power station in all photovoltaic power stations which are located in the same region grid as a target power station and achieve the remote monitoring; taking the power generation parameter of the benchmarking power station as a reference, estimating to obtain a power generation state theoretical value of the target power station through a preset algorithm, and comparing the power generation state theoretical value with a powergeneration state actual value obtained by relying on the state grid new energy cloud, so that on the premise of not additionally adding any hardware equipment, the power generation state of the target power station is effectively evaluated, the problem that the power generation management level of the power station is low due to the fact that most small and medium-sized household photovoltaic power generation systems lack remote monitoring is solved, and the implementation cost is low.

Distributed power supply scheduling control method based on edge calculation

Abstract: The invention relates to the technical field of power distribution networks, in particular to a distributed power supply scheduling control method based on edge calculation. The method comprises the following steps: A) obtaining a topological structure and scheduling data of a target distribution network, and listing a transformer, energy storage equipment and a wind-solar power station in the target distribution network; B) reading an active power prediction value and a reactive power prediction value of the load in the next time period, and predicting the output of the wind-solar power station; C) calculating a characteristic value, and if the characteristic value is smaller than a set threshold value, entering a step E); D) only outputting active power by the wind-solar power station, and returning to the step B); E) calculating the ratio of active power to reactive power output by the wind-solar power station; and F) respectively outputting the real-time output of the wind-solar power station as active power and reactive power according to the ratio, and returning to the step B. The method has the substantive effects of reducing the influence on the power distribution network when the renewable energy power station with unstable output is accessed to the power distribution network, being beneficial to expanding the access of renewable energy, and being suitable for the power distribution network with high permeability.

Random forest-based method for photovoltaic systems based maximum power point tracking

Abstract: Solar energy is one of the key sources of green energy because it is free of emissions, pure, limitless, and plentiful. Solar power has gained tremendous interest across the globe, and one of its key duties is to produce as much solar power as possible in varying weather conditions. Photovoltaic generation’s intermittent and unregulated characteristics have a major effect on power stability. To minimize these factors, the predictability of photovoltaic generation needs to be increased. As photovoltaics are increasingly evolving, the share of PV generation in the electricity trading industry is rising. However, the model’s accuracy is often low in conventional modeling because of unnecessary initial data noise or inappropriate parameter configuration. The key factor analysis in this article and the K-means algorithm are clustered and the random forest (RF) algorithm. Core components and K-means are used to get the hourly points’ characteristics close to the projected time points, and then the input data are filtered to mitigate interaction with the noise data.