V. Prema

Predictive models for power management of a hybrid microgrid — A review

The energy demand of the human civilization is increasing day by day, which has made man to look for alternative sources, considering that the fossil fuels, which are the principal sources of energy, are depleting. In addition, the carbon footprint left over by fossil fuels has a detrimental effect on the earths environment. This has entailed researchers to focus their attention to environment friendly and renewable energy sources. The primary considerations in this subject viz. solar and wind energy sources pose their own challenges to researchers, an essential component being their stochastic nature. Weather conditions, weather patterns and the site chosen have a direct impact on the effectiveness of the implemented system. A helping hand is extended by nature with the fact that the availability of these two sources is complimentary to one another, assuring a power source in all weather conditions. This has paved way for researchers to converge their studies on Hybrid power systems which employ multiple types of power generators to cater to the demand. A Predictive Power Management scheme which incorporates a forecast of the power generation capability of each generator, the load demand and other site-specific parameters is vital to extract the best of the implemented system. Such a management system, which makes a long term forecast, minimizing errors on the behavioral patterns of wind and solar energy has become a major subject matter of study for researchers across the globe. This paper gives an overview of the power management strategies. Different predictive power management topologies, advantages and challenges are discussed.

An effective dispatch strategy for hybrid power management

Decentralised power generation systems based on renewable energy can play an important role in hastening electrification of isolated areas in developing countries. Next generation, smart microgrids present sophisticated solutions for monitoring and visualising of all available resources and calculating of optimised schedules for dispatch of resources to ensure the most cost effective and sustainable operational strategy. This paper discusses an algorithm for economic dispatch of an off-grid microgrid comprising of a PV-array and wind turbine combined with a battery and variable speed diesel generator backup. The nature of the dispatch strategy developed has been compared to existing dispatch strategies. The sharing of the load amongst the various sources has been simulated for the given setup and the results are analysed.

Multivariate regression models for prediction of wind speed

As we progress in both time and technology, our energy needs are rising at an exponential level and hence we need to tap unconventional sources of energy more efficiently. Wind energy is one such source and this paper presents a method to predict the speed of wind, on which the wind energy generated, depends more efficiently and hence avoid both costly overproduction and underproduction. This can be achieved by statistical methods wherein data in large numbers are collected, analyzed for functional relationship using Multivariate regression models. The results obtained are then compared with the actual values available for validation.

Multivariate regression for prediction of solar irradiance

This paper describes regression models to forecast solar irradiance for a short term (or period). The regression models enable the prediction of solar irradiance in minute values over a period of a few days. A single variate regression model is used and various plots obtained between solar irradiance as dependent variable and air temperature and relative humidity as independent variables have been studied. Optimal range for prediction using regression is decided. To obtain accuracy multivariate regression is carried out It also presents new multifunctional relationship between solar irradiance, air temperature and relative humidity. This multifunctional regression relationship gives more accurate results compared to other methods having single variable. In this regression model solar irradiance follows an increasing trend upto a particular temperature after which it shows decreasing trend and hence it has been modeled with three equations.

Multivariate wind power forecast using artificial neural network

Power generations from renewable sources of energy like solar and wind is catching up rapidly. There is a dire need for forecasting the generation in order to have better load scheduling as the generation is uncertain because the weather is erratic and the generation depends on a lot of factors. Therefore with greater penetration of renewable sources in power generation, the focus is shifting towards generation forecasting. This paper proposes predictive models for wind power generation using non-linear auto regressive neural network. Three multivariate models are developed for a day ahead prediction of wind power generation. A comparative study is done on the errors and it is found that wind speed is highly dependent on wind direction. A model with wind speed and wind direction as inputs gives better prediction.

Efficient MPPT for a stand-alone photovoltaic system

This paper presents an efficient MPPT controller. Incremental conductance algorithm is chosen for MPPT to improve the efficiency. The advantage of this algorithm is that the operating point does not oscillate around the Maximum Power Point. The algorithm also detects the atmospheric conditions. The operating point gets shifted according to the irradiance. The DC-DC converter chosen is a MOSFET based Boost converter as the Maximum Power Point is efficient in boost converter. The digital controller used is a DSP controller. The design and Simulations of MPPT are done using the MATLAB IDE link to DSP TMS320F2812. A comparison of MPPT control is done with the power control method known as limit cycle power control or sequential switching regulator. The simulation models are built using MATLAB for both the control methods and the results are compared.

Application of Hybrid Neuro-Wavelet Models for Effective Prediction of Wind Speed

Severe energy crisis and depletion of fossil fuels necessitates more number of installations of wind farms. Accurate wind forecast is crucial in the efficient utilization and power management of wind farms connected to a grid or in conjunction with other sources such as solar, DG, battery, etc. This paper proposes a hybrid neuro-wavelet predictive tool to predict wind speed which combines the advantages of both wavelet decomposition and neural network. Wavelet decomposition is used to filter out the high frequency outliers in the wind speed, thus making a smooth data to make the prediction accurate. The filtered data is used to train the neural network. Four different models are proposed. NAR-TS model and NAR-Wavelet models are univariate models with past values of wind speed as input. In NAR-TS model time series values are directly applied as input to neural network, whereas in NAR-Wavelet model input to the neural network is the wavelet decomposed data. In a similar way NARX-TS and NARX-Wavelet models are developed with multivariate neural network, where the inputs are air temperature, relative humidity and wind speed which is the feed back. Each of these models are used to predict 4.5 hours ahead and 18 hours ahead predictions. The Mean Average Percentage Error (MAPE) values are calculated for each model and the results are compared.

Effective battery usage strategies for hybrid power management

In developing countries there are a large number of isolated areas not connected to the main grid. The energy demands of these remote, isolated communities are met by diesel operated power systems. These systems are often highly polluting and with time, an effort is made to make a transition to systems wherein single or multiple renewable energy sources are used to augment the diesel generators and thereby reduce the environmental footprint. Majority of such off-grid hybrid systems include a massive battery bank to store excess energy to supply the user when the renewable is inefficient. Inclusion of such large batteries results in additional investment costs and recycling issues. Therefore it is necessary to minimise the battery size as far as possible, and operate the system under a strategy to ensure maximum utilisation of renewable supplies and minimal DG usage. In this paper the power management strategy for a PV-battery and VSDG system was developed. A battery was modelled in MATLAB and four different battery operation algorithms were simulated for a set of data comprising of load and solar power output for every ten minute period for a day and the effective usage of DG was compared and analysed.

Sizing of microgrids for Indian systems using HOMER

The energy needs of mankind are rising at an exponential level and it necessitates tapping of unconventional sources of energy more efficiently. In a country like India where around 300 million people lack electricity, the concept of micro grid plays a significant role. Hybrid systems with wind and solar energy are ideal for village electrification. The major challenge in designing hybrid system is the optimization of the various sources involved in hybrid systems. This paper gives a perfect model of a micro grid with optimized size of wind, solar, diesel generator and battery using HOMER software. A detailed cost analysis is carried out and the designed system is compared with only DG system in terms of emission, fuel consumption and cost of the system over the next 15 years.

Novel training strategies for wavelet-neuro models for wind speed prediction

The wind energy provides opportunities to generate power cheaply and cleanly without affecting the environment. The problem with wind energy is its variable and intermittent nature. Thus a large-scale introduction of wind power causes a number of challenges for the electricity market and power system operators who need to deal with the variability and uncertainty in wind power generation when making their scheduling and dispatch decisions. Accurate wind power forecast can solve these problems to a great extent. This paper proposes three novel strategies to train neural network to improve the prediction accuracy. Wavelet decomposition is used to filter out the high frequency outliers in the wind speed, thus making a smooth data to make the prediction accurate. The filtered data is used to train the neural network. In recursive training, the number of prediction steps during the training process, are reduced to increase the prediction accuracy. The neural network is re-trained with these predicted values. In conditional training, a pre-determined threshold level is set for the error. The training stops when the error falls below this level. In parallel training, 10 parallel networks is created with either recursive or conditional training, each of which is trained separately and the final predicted wind speed is the mean of the prediction done by individual parallel path.