This paper describes a new approach to online forecasting of power production from PV systems. The method is suited to online forecasting in many applications and in this paper it is used to predict hourly values of solar power for horizons of up to 36 h. The data used is 15-min observations of solar power from 21 PV systems located on rooftops in a small village in Denmark. The suggested method is a two-stage method where first a statistical normalization of the solar power is obtained using a clear sky model. The clear sky model is found using statistical smoothing techniques. Then forecasts of the normalized solar power are calculated using adaptive linear time series models. Both autoregressive (AR) and AR with exogenous input (ARX) models are evaluated, where the latter takes numerical weather predictions (NWPs) as input. The results indicate that for forecasts up to 2 h ahead the most important input is the available observations of solar power, while for longer horizons NWPs are the most important input. A root mean square error improvement of around 35% is achieved by the ARX model compared to a proposed reference model.

/pdf/online-short-term-solar-power-forecasting-1ljgpbshkb.pdf

Online Short-term Solar Power Forecasting

Integration of photovoltaics into power grids is difficult as solar energy is highly dependent on climate and geography; often fluctuating erratically. This causes penetrations and voltage surges, system instability, inefficient utilities planning and financial loss. Forecast models can help; however, time stamp, forecast horizon, input correlation analysis, data pre and post-processing, weather classification, network optimization, uncertainty quantification and performance evaluations need consideration. Thus, contemporary forecasting techniques are reviewed and evaluated. Input correlational analyses reveal that solar irradiance is most correlated with Photovoltaic output, and so, weather classification and cloud motion study are crucial. Moreover, the best data cleansing processes: normalization and wavelet transforms, and augmentation using generative adversarial network are recommended for network training and forecasting. Furthermore, optimization of inputs and network parameters, using genetic algorithm and particle swarm optimization, is emphasized. Next, established performance evaluation metrics MAE, RMSE and MAPE are discussed, with suggestions for including economic utility metrics. Subsequently, modelling approaches are critiqued, objectively compared and categorized into physical, statistical, artificial intelligence, ensemble and hybrid approaches. It is determined that ensembles of artificial neural networks are best for forecasting short term photovoltaic power forecast and online sequential extreme learning machine superb for adaptive networks; while Bootstrap technique optimum for estimating uncertainty. Additionally, convolutional neural network is found to excel in eliciting a model's deep underlying non-linear input-output relationships. The conclusions drawn impart fresh insights in photovoltaic power forecast initiatives, especially in the use of hybrid artificial neural networks and evolutionary algorithms.

A review and evaluation of the state-of-the-art in PV solar power forecasting:Techniques and optimization

Thyristor-Based FACTS Controllers for Electrical Transmission Systems

https://isiarticles.com/bundles/Article/pre/pdf/46311.pdf

A review on interval type-2 fuzzy logic applications in intelligent control

Photovoltaic power generation forecasting is an important topic in the field of sustainable power system design, energy conversion management, and smart grid construction. Difficulties arise while the generated PV power is usually unstable due to the variability of solar irradiance, temperature, and other meteorological factors. In this paper, a hybrid ensemble deep learning framework is proposed to forecast short-term photovoltaic power generation in a time series manner. Two LSTM neural networks are employed working on temperature and power outputs forecasting, respectively. The forecasting results are flattened and combined with a fully connected layer to enhance forecasting accuracy. Moreover, we adopted the attention mechanism for the two LSTM neural networks to adaptively focus on input features that are more significant in forecasting. Comprehensive experiments are conducted with recently collected real-world photovoltaic power generation datasets. Three error metrics were adopted to compare the forecasting results produced by attention LSTM model with state-of-art methods, including the persistent model, the auto-regressive integrated moving average model with exogenous variable (ARIMAX), multi-layer perceptron (MLP), and the traditional LSTM model in all four seasons and various forecasting horizons to show the effectiveness and robustness of the proposed method.

Short-term photovoltaic power forecasting based on long short term memory neural network and attention mechanism

Solar photovoltaic power forecasting using optimized modified extreme learning machine technique

A comparative study on short-term PV power forecasting using decomposition based optimized extreme learning machine algorithm

In present scenario the energy system face various challenges as the demand for energy is increasing significantly and the resources in terms of fossil fuels are limited, need for renewable resources have became very much vital at present. Accurate and reliable solar power forecasting is essential for the legitimate functioning of the power system. Given momentous uncertainties involved in solar power generation due to variation of temperature and irradiance, forecasting provides a unique solution for the uncertainties and variability's in solar data. In this paper a forecasting method has been mentioned that is contingent on a hybrid empirical mode decomposition (EMD) and Extreme Learning Machine (ELM). The non stationary time series is further decomposed into distinct intrinsic mode functions (IMF). A short term forecasting is also carried out in this work to prove the accuracy of the given model. This model is implemented in MATLAB/SCRIPT environment.

Solar power forecasting using a hybrid EMD-ELM method

Photovoltaic (PV) system uses maximum power point tracking (MPPT) technique to maximize its power output. This paper presents a swarm inspired optimization based control method for the MPPT of a PV system. In particular, control methods of perturb and observe (P&O)-proportional integral (PI) and optimisation of this controllers using spider monkey optimisation (SMO) for MPPT control of PV generator are proposed. A well known PI controller is tuned using spider monkey algorithm and tested on a simulated stand alone PV system with local load. The proposed SMO based PI controller enhances the P&O MPPT technique to track the maximum power point (MPP) more rapidly and accurately. The design algorithm of this controller is presented with its simulation results. Simulation results under different environmental conditions are presented and discussed to verify the satisfactory performance of the proposed control, in which the optimised P&O-PI controller by SMO algorithms gives the better performance technique over the traditional control methods.

Spider monkey based improve P&O MPPT controller for photovoltaic generation system

This paper presents the design of a robust nonlinear controller for a parallel AC-DC power system using a Lyapunov function-based sliding mode control (LYPSMC) strategy. The inputs for the proposed control scheme are the DC voltage and reactive power errors at the converter station and the active and reactive power errors at the inverter station of the voltage-source converter-based high voltage direct current transmission (VSC-HVDC) link. The stability and robust tracking of the system parameters are ensured by applying the Lyapunov direct method. Also the gains of the sliding mode control (SMC) are made adaptive using the stability conditions of the Lyapunov function. The proposed control strategy offers invariant stability to a class of systems having modeling uncertainties due to parameter changes and exogenous inputs. Comprehensive computer simulations are carried out to verify the proposed control scheme under several system disturbances like changes in short-circuit ratio, converter parametric changes, and faults on the converter and inverter buses for single generating system connected to the power grid in a single machine infinite-bus AC-DC network and also for a 3-machine two-area power system. Furthermore, a second order super twisting sliding mode control scheme has been presented in this paper that provides a higher degree of nonlinearity than the LYPSMC and damps faster the converter and inverter voltage and power oscillations.

Niranjan Nayak

Papers

Solar photovoltaic power forecasting using optimized modified extreme learning machine technique

A comparative study on short-term PV power forecasting using decomposition based optimized extreme learning machine algorithm

Solar power forecasting using a hybrid EMD-ELM method

Spider monkey based improve P&O MPPT controller for photovoltaic generation system

Nonlinear control of voltage source converters in AC-DC power system