Search

Wind power forecasting, to a certain extent, will transform the random fluctuation of wind power output into a known situation, which is one of the effective approaches to deal with large-scale wind power integrated into power grid. Due to the use of only historical data and the lack of new information, the accuracy of ultra-short-term wind power prediction (WPP) is still not satisfactory. Therefore, a combined prediction method based on the day-ahead Numerical Weather Prediction (NWP) location technology is proposed. Firstly, the time points with low forecasting accuracy of rolling WPP are approximately located by the NWP information and time windows, and then the hybrid approach combined with neural network and persistence method is presented to predict the future wind power output. The results of the case study show that compared with other classical prediction methods, this method can effectively improve the ultra-short-term prediction accuracy of wind power and verify the effectiveness of the proposed method.

The short-term probabilistic prediction of wind power has the characteristics of spatial dependence and time-series dependence. Considering the two characteristics at the same time can improve the prediction level. In this paper, a probabilistic short-term wind power prediction model considering the temporal and spatial dependence of prediction error is proposed. Considering the coupling relationship between the two properties, the NWP(Numerical Weather Prediction) wind speed point prediction error in the historical period is hierarchical clustering, and the empirical distribution model is used to fit the probability distribution of the error under different wind conditions; the cumulative empirical distribution probability value corresponding to the NWP wind speed at the time to be predicted is bootstrap sampling; under the given confidence level, the possible wind speed at each time point to be predicted in the short term is calculated The power fluctuation range of the generator. The test results show that this method can ensure the statistical significance and fitting stability of the sub sample set at the same time, and improve the classification accuracy of the days to be predicted. Compared with considering a single property, the result of probability prediction performs better on multiple evaluation indexes.

In view of the influence of the randomness, volatility and anti-peak-regulation characteristics of large-scale grid-connected wind power output on the grid’s peak-regulation and dispatching, the technical means of joint operation of concentrated Solar power system with thermal storage and wind power is adopted. The complementarity of solar-thermal and wind power output and the dispatchability of thermal storage are used to suppress the fluctuation of wind power output, reduce the peak-valley difference of power grid load, reduce the cost of thermal power peak shaving, and reduce wind power abandonment. Considering that the uncertainties of wind power and light power bring about the increase of dispatch operation risk, the conditional value-at-risk (CVaR) is used to measure the risk loss caused by uncertainty factors to dispatch operation. Aiming at the lowest total dispatch cost and risk cost, the economic dispatch strategy of solar thermal- wind power system considering CVaR is studied. Finally, the feasibility and rationality of the dispatch strategy and the model are verified by a simulation example.

With the development of new energy industry, it’s beneficial to analyze the characteristics of new energy which stabilize the operation of the power grid and economic dispatch. It’s more and more important to model the uncertainty of new energy. Although there are some probabilistic modeling methods based on real load data for wind power output, they are not accurate and have high computational complexity. In order to solve these problems, this paper proposes a source-charge probability distribution model based on variational autoencoder. The model uses deep learning techniques to perform variational autoencoder and probabilistic modeling for real-time wind power data. This unsupervised method can learn the characteristics of wind power data and generate new data based on the characteristics of the observation without the scene reduction process. It improves the efficiency and quality of scene generation.

With the development of new energy, such as wind energy, the scale of wind power installation is gradually increasing. It causes some areas rich in wind resources cannot fully utilize and consume wind power, and then the wind abandonment problem is particularly prominent. At the same time, because of the incomplete basic operation data collection and the complex operation characteristics under different time scales, the matching degree of new energy and multi-load, composed of controllable load and energy storage, is not high at different time scales of power and electricity. In order to solve the problems above, this paper proposes a source-load value matching method for many uncertain factors (such as weather, wind speed, etc.), which is to match the source and load value by calculating the matching degree. The method can be used to improve the nearby wind power accommodation capacity of new energy, so as to improve the utilization rate of new energy. So the measure in this paper is the matching degree. The greater the matching degree is, the better the capacity of wind power accommodation is. In this paper, six schemes are designed and experiments are carried out to verify the reliability and practicability of the method. The scheme with largest matching degree is the best regional scheme of wind power accommodation.

Wind power consumption is often curtailed by the inflexible operation of combined heat and power (CHP) units due to the strong coupling relationship between power generation and heating supply. Besides, the time delay of district heating network and indoor temperature usually affects the operation plan of the district heating system (DHS) heavily. Hence, a practical CHP-DHS model and multiregional coordinated operation strategy based on model predictive control are developed for planning and operating this CHP system. The thermal characteristics of demand side such as the thermal inertia of buildings and thermal comfort are taken into consideration. Three types of heat source are considered: CHP units, electrical boilers, and heat storage tanks. The objective of the optimization is to minimize the overall generation costs of CHP units and conventional thermal power units. Part of real provincial power system and a municipal DHS in the Northeastern China are used in this case study. The results show that the proposed method has better performance in wind power integration and indoor temperature control.

Combined cooling, heating and power (CCHP) microgrid has the advantage of high energy utilization efficiency. The fluctuation of renewable energy sources and multiple load demands challenges the economic operation of CCHP microgrid. In this paper, we propose a novel two-stage coordinated control approach for CCHP microgrid energy management, which consists of two stages: the economic dispatching stage (EDS) and the real-time adjusting stage (RTAS). In EDS, it utilizes a model predictive control incorporating piecewise linear efficiency curves to schedule the operation based on the forecast information. In RTAS, the schedule obtained in EDS is adjusted based on the real-time information to tackle the power fluctuations. A typical-structure CCHP microgrid is analyzed in the case study and simulation results are presented to demonstrate the performance of the proposed two-stage coordinated control approach.

In this paper, the uncertainty of diesel generator (hereinafter referred to as DG) output and load demand is studied by multi scenario analysis. Firstly, this paper studies the timing characteristics of DG power generation and load demand, and describes the annual and daily distribution of photovoltaic power generation, wind power generation and load demand. Then it introduces the method of multi scene analysis, describes the process of dealing with the uncertainty in multi scene in detail, and generates enough scenes through the density function of classification probability to get the probability of each scene. Finally, H-K compound clustering algorithm is used to solve the above problems. The scale scene is compressed to get a typical “planning scene”.

A lot of mechanical parts are subject to failure due to the deterioration. Usually the preventive maintenance is taken to ensure the safety and reliability. Therefore, it is very important to study the gradual reliability design of the mechanical part for improving the gradual reliability of the mechanical system under the condition of considering the preventive maintenance. Beta distribution is employed to describe the randomness of the mechanical part state after the preventive maintenance. The deterioration process of the mechanical part is modeled using the nonstationary Gamma process. The gradual reliability model before the first preventive maintenance is proposed according to the gradual failure principle and using the initial state distribution and the properties of Gamma process. Then, the gradual reliability model between any two times of preventive maintenance is also derived. Subsequently, the sensitivity equations of the proposed gradual reliability model to each parameter are given. The application process and practicality of the proposed approach are described by a numerical example. This work solved the problem where the maintenance has not been well considered in the reliability design of the mechanical part and contributed to the theory and method of improving the safety and reliability operation of the mechanical system.