Which algorithm is best for energy management in smart grid?
To work on 3 objectives, like
1. operating cost
2. pollution emissions
3. availability of renewable energy energy sources.
Which algorithm (must new) can we use?
First, you've two options: you can combine all of your three objectives into a single objective and solve it (weight-sum method, goal programming etc) or you could use pareto-based methods to solve multi-objectives.
Secondly, there is not any single optimization algorithm which can outperform all other algorithms as the result depends on the problem size, hyperparameters of algorithms etc. So, commonly multiple algorithms are used to solve the same optimization problem for many runs (typically 30) and the algorithm which gives the average minimum value (for minimization) is chosen as the best algorithm for that problem.
However, algorithms such as PSO, GA, Ant-Colony are well used for single objective problem and NSGA-I,II,III are very popular for the multi-objectives optimization problem.
Recently, reinforcement learning is getting attention in this regard.
Dear Kalim:
You can benefit from this valuable article (2020)..
Tilte:
"State-of-the-Art Artificial Intelligence Techniques for Distributed Smart Grids: A Review".
Authors:
Syed Saqib Ali and Bong Jun Choi *
School of Computer Science and Engineering, Soongsil University, Seoul.
Subtract:
The power system worldwide is going through a revolutionary transformation due to the integration with various distributed components, including advanced metering infrastructure, communication infrastructure, distributed energy resources, and electric vehicles, to improve the reliability, energy efficiency, management, and security of the future power system. These components
are becoming more tightly integrated with IoT. They are expected to generate a vast amount of data to support various applications in the smart grid, such as distributed energy management,generation forecasting, grid health monitoring, fault detection, home energy management.
With these new components and information, artificial intelligence techniques can be applied to automate and further improve the performance of the smart grid. In this paper, we provide a comprehensive review of the state-of-the-art artificial intelligence techniques to support various applications in a distributed smart grid. In particular, we discuss how artificial techniques are applied to support the integration of renewable energy resources, the integration of energy storage systems,
demand response, management of the grid and home energy, and security. As the smart grid involves various actors, such as energy produces, markets, and consumers, we also discuss how artificial intelligence and market liberalization can potentially help to increase the overall social welfare of the grid. Finally, we provide further research challenges for large-scale integration and orchestration of automated distributed devices to realize a truly smart grid.
Also this article (2020):
"Deep Reinforcement Learning for Smart Building Energy Management: A Survey"
Liang Yu, Member, IEEE, Shuqi Qin, Meng Zhang, Chao Shen, Senior Member, IEEE,
Tao Jiang, Fellow, IEEE, Xiaohong Guan, Fellow, IEEE.
Abstract:
Global buildings consumed 30% of total energy and generated 28% of total carbon emission in 2018, which leads to economic
and environmental concerns. Therefore, it is of great significance to reduce energy consumption, energy cost and carbon emission
of buildings while maintaining user comfort. To this end, several challenges have to be addressed. Firstly, it is very challenging
to develop a building thermal dynamics model that is both accurate and efficient enough for building control. Secondly, there are
many kinds of uncertainties. Thirdly, there are many spatially and temporally operational constraints. Fourthly, building energy
optimization problems may have extremely large solution spaces, which can not be solved in real-time by traditional methods.
Fifthly, traditional building energy management methods have respective applicable premises, which means that they have low versatility when confronted with varying building environments. As a general artificial intelligence technology, deep reinforcement learning (DRL) has the potential of addressing the above challenges. Thus, this paper presents a comprehensive literature review on DRL for smart building energy management (SBEM). To be specific, we first introduce the fundamentals of DRL and provide the classification of DRL methods used in existing works related to SBEM. Then, we review the applications of DRL in a single building energy subsystem, multiple energy subsystems of buildings, and building microgrids, respectively. Furthermore, we identify the unsolved issues and point out the possible research directions of applying DRL. Finally, we summarize the lessons learned from this survey.
With my best wishes..
I hope it will be helpful ...
Flower pollination and Mine Blast for optimizing multi objective functions.
I suggest MCDM techniques in combination with commercial software such as HOMER, PVSYST, TSOL, and etc.
Dear Mehdi Jahangiri sir,
I am completely agree with you, MCDM combined with such technical tools i.e., PVSYST or HOMER may provide more roust results. MCDM has capacity to analyze non-technical parameters precisely, whereas, PVSYST (on-grid) or HOMER (off-grid) may analyze technical parameters.
Article CO2 Capture, Use, and Storage in the Cement Industry: State ...
Also this article can help u.
I think the best algorithms for energy management are ANN, PSO, and multi-agent system
Ant colony Optimizzation algorithm would be good . You can give a try.
Kalim Ullah
IT would depend on what type of objective u have made linear or nonlinear and in the context of that u r constraints will matter whether u have inequality constraint or equality constraint. depending on all you need to find which type of solver u will use .
Mehdi Jahangiri
Dear mehdi please, can you explain what do you mean MCDM?
Dear Seydali Ferahtia
Please see the below paper:
Using fuzzy MCDM technique to finding the best location in Qatar for exploiting wind and solar energy to generate hydrogen and electricity
Dear sir
The algorithm is based on your application is it in isolated DC Micro grid or on grid mode and also it depends on the types of the renewal energy sources
I have tried PSO, GA, DE, and HOMER. In my, problem DE provide better result but these results vary from types of source, number of sources, constraints. so maybe another algorithm also be best its vary from problem to problem
A fast randomized first-order optimization algorithm or You can try Adaptive Gradient Descent (i.e. A method of stochastic optimization-ADAM)
https://machinelearningmastery.com/adam-optimization-algorithm-for-deep-learning/
Thanks for the question. As per no free lunch theorem no algorithm is best suited for all types of objective functions i.e., linear, non-linear and integer. Based on the objective function and constraints the structure of search space changes. Further, the presence of constraints generates discontinuities in the solution space. Therefore, solve a set of algorithms and select the best one which is giving trade-off between above issues for the objective function chosen.
Based on the objective function and constraints the structure of search space changes.
https://ieeexplore.ieee.org/document/7350599
It depends on your goal. Sometimes for engineering problems rather than the general algorithm the solution to the particular problem would be the best. Furthermore, operating cost general would have correlation with pollution produced by energy generator. So that constraint could be dropped. Now depending on the situation direct logic specifically designed for the task, has the potential to provide similar solution to optimisation techniques. The result would be suboptimal, however within the errors of pollutants generator predictions and the renewable energy generation prediction, making more accurate solution does not make much sense in cloudy regions, as clouds could easily introduce 90% error to prediction. Furthermore, the speed of calculation for logical approaches could be thousands times faster than others. That makes standard techniques of optimisation not really applicable to the engineering problems for the needed time intervals. You might try these to adapt these one to your problem https://www.sciencedirect.com/science/article/abs/pii/S0959652621030419 or design your own, by that means it would be also new, just make sure that mathematically it has convergence to some set of bounded values. In general specific problems requires specific solution depending on the outcome one seek.
