An Optimized and Decentralized Energy Provision System for Smart Cities

Round 1

Reviewer 1 Report

Dear Authors,

The article deals with the subject of smart city networks, the smart city energy model, prosumer communities producing, consuming and sharing clean energy, as well as the use of modern optimization methods (GA, PSO) in the given topics. However, the area of ​​renewable energy is not the main topic of the article. The main issues discussed in the paper are: Wireless Sensor Network, Advanced Metering Infrastructure and blockchain technology. For this reason, I consider that this work is not suitable for publication in the Energies journal. The authors should looking for another journal related to the keywords indicated by them. Additionally, on the basis of the implemented input data of the analyzed example, is impossible to refer to the correctness of the obtained results. The information in the article cannot be used to recreate the research described, which should be one of the elements of scientific papers. Other researchers should be able to recreate conducted research, making their own conclusions, and conducting a scientific discussion.

Author Response

The Introduction and implementation of methods sections have been modified to present a better understanding of the concept and the errors in the paper have been corrected as suggested by the respected reviewer.

Reviewer 2 Report

This paper proposed the application of bio-inspired algorithms in order to improve routing process by finding the shortest path for traversing the whole network and enhancing the system performance in terms of efficient selection of cluster head, reduced energy consumption, and extended network lifetime. The paper has a proper quality and can be published after applying the below modifications:

1. Firstly, please list the main challenges in this field, and after that, bold the contributions of your work in the last paragraph of the introduction.
2. The literature review is comprehensive and there is a deep history of the literature, but please consider more the application of both PSO and GA references. Meanwhile, that would be great to have a look at the relevant published papers in the Energies- journal as well.
3. In Figure 2, it would be better to change the "group's best location" to "global best location".
4. After formulas line, the capital character of "Where" should be corrected to "where". and in the PSO formula, the 'w' is the inertia weight, please correct it.
5. In Figure 10, it is not clear which optimisation method is applied PSO or GA? the name of the method should be added in the caption. 
6. How you can compare PSO and GA performance when the number of evaluation is different? please provide a fair comparison situation means the same population size and number of generation.

 

 

 

Author Response

This paper proposed the application of bio-inspired algorithms in order to improve routing process by finding the shortest path for traversing the whole network and enhancing the system performance in terms of efficient selection of cluster head, reduced energy consumption, and extended network lifetime. The paper has a proper quality and can be published after applying the below modifications:

Comment 1: Firstly, please list the main challenges in this field, and after that, bold the contributions of your work in the last paragraph of the introduction.

Response: As per the instruction of the respected reviewer, we have improved the introduction part by highlighting the challenges in the field of wireless sensor networks and centralized energy systems and added our contributions at the end of the introduction.

Comment 2: The literature review is comprehensive and there is a deep history of the literature, but please consider more the application of both PSO and GA references. Meanwhile, that would be great to have a look at the relevant published papers in the Energies- journal as well.

Response: Thank you for this suggestion. We have improved the literature review of this paper by adding few more applications of PSO and GA algorithms.

Comment 3: In Figure 2, it would be better to change the "group's best location" to "global best location".

Response:  We have modified the Figure. We apologize for this mistake.

Comment 4: After formulas line, the capital character of "Where" should be corrected to "where". and in the PSO formula, the 'w' is the inertia weight, please correct it.

Response: Thank you for pointing this out. The error has been corrected. We apologize for this mistake.

Comment 5: In Figure 10, it is not clear which optimisation method is applied PSO or GA? the name of the method should be added in the caption.

Response: In Figure 10, the optimization method used for calculating the residual energy in every round of iteration is PSO algorithm. In each round, the node having the highest residual energy is elected as the cluster head (as shown in Figure 9). The title of Figure 10 has been modified now in the paper.

Comment 6: How you can compare PSO and GA performance when the number of evaluation is different? please provide a fair comparison situation means the same population size and number of generation.

Response: We have used PSO in selecting the location of cluster head and GA for routing. Since both the algorithms are used in completely different scenarios, their population size and number of generations are chosen accordingly. Had they been used in same application, we could have compared both the algorithms. But this paper focuses mainly on how these two optimization techniques can be used in the above two applications in the field of WSN.

Reviewer 3 Report

• The title of the article does not follow the proposal in this study. The authors must change the title and they can propose for example a web application to the service of a smart city.
• What is the energy dispatching in figure 5 between the different users?
• What are the storage elements and the storage capacity in the smart energy community model?
• What type of genetic algorithm was used in this study? it should be compared with other types of genetic algorithms such as NSGA II or III to show the effectiveness of the proposed method.
• On what basis was the cost function found?
• Figure 15 needs to be explained further, is this application in real time? 
• All the equations must be numbered.
• The simulation results are poor.
• The references must be completed with other references.

The authors can use the two references below to improve their paper.

• A smart cyber physical multi-source energy system for an electric vehicle prototype, ELSEVIER 2020.
• Grid of Hybrid AC/DC Microgrids: A New Paradigm for Smart City of Tomorrow, 2020 IEEE 15th International Conference of System of Systems Engineering (SoSE).

Author Response

Comment 1: The title of the article does not follow the proposal in this study. The authors must change the title and they can propose for example a web application to the service of a smart city.

Response: Thank you for this suggestion, but our project is not limited to a blockchain based web application. This web application is a part of the optimized and decentralized energy provision system. Our paper basically presents a smart city energy model in which houses are grouped into energy communities which consists of prosumers, local consumers and a smart microgrid. For intelligent management of microgrids in this interconnected system, wireless sensor networks are extensively used and therefore their optimization is necessary for conserving energy and increasing network lifetime. For this purpose, GA and PSO algorithms are proposed and implemented. In this model, blockchain technology is used to protect the network privacy along with facilitating a peer to peer energy trading between prosumers and consumers with the help of advanced metering infrastructure. This decentralized trading system is established using a blockchain based web application which is implemented in this paper. We have modified both introduction and implementations part of this paper to make the proposal in this study follow the title of this article.

Comment 2: What is the energy dispatching in figure 5 between the different users?

Response: Energy is generated independently by the prosumers and it is sold to other prosumers and local consumers depending on the demand. Figure 5 (now changed to Figure 4) has been modified by adding flow of energy in the diagram.

Comment 3: What are the storage elements and the storage capacity in the smart energy community model?

Response: These are decentralized storage systems that are coupled with local renewable energy generations of each house. The storage capacity is less in decentralized storage systems since there is no large scale production of electricity. The same has been added in the paper.

Comment 4: What type of genetic algorithm was used in this study? It should be compared with other types of genetic algorithms such as NSGA II or III to show the effectiveness of the proposed method.

Response: We have used a modified genetic algorithm with enhanced crossover and mutation operators.

Comment 5: On what basis was the cost function found?

Response: The cost function should be a sum of average Euclidean distance of a node from all other member nodes, its distance from the base station and its residual energy, since all three factors are necessary to select the cluster head node which should be closer to all member nodes and base station and must have highest residual energy among other nodes, as per the previous research works. We have tested our PSO code with different values of alpha, beta and gamma. Based on different simulation results, we have selected the values of alpha, beta and gamma as 0.38, 0.38 and 0.18 respectively, in order to get the best results. We have updated the paper by mentioning these values of the constants in the paper.

 

Comment 6: Figure 15 needs to be explained further, is this application in real time? 

Response: Yes, this is a real time application. Every time a prosumer sells energy units on the web application, a new record is created having details of the prosumers like the public key and price of energy unit to be sold assuming the cost of 1 kwh of electricity is 0.14 USD or 0.00026 ETH. In our model of interconnected prosumer-consumer network, consumers can buy energy directly from a prosumer through blockchain. Here, a consumer can pay Ethers online using our web app by clicking on the ‘Buy’ button that will generate a MetaMask confirmation first for signing the transaction and then create a block that will be added to the Ganache blockchain after the transaction is validated through mining, all in real time. Ethers will be transferred from buyer’s account to seller’s account during the transaction and the same will reflect in their MetaMask wallet balance. Here, we have used a private blockchain Ganache which behaves same as a real Ethereum blockchain (like the one in etherscan.io) but it will not cost us any real money for making transactions.

 

Comment 7: All the equations must be numbered.

Response: Thank you for pointing this out. The equations have been numbered.

Comment 8: The simulation results are poor.

Response: Authors made a sincere attempt to improve the simulation results in the revised manuscript.

Comment 9: The references must be completed with other references.

Response: We have added few more references in the paper along with the suggested references.

Comment 10: The authors can use the two references below to improve their paper.

• A smart cyber physical multi-source energy system for an electric vehicle prototype, ELSEVIER 2020.
• Grid of Hybrid AC/DC Microgrids: A New Paradigm for Smart City of Tomorrow, 2020 IEEE 15th International Conference of System of Systems Engineering (SoSE).

Response: Thank you for this suggestion. We have added these references in our paper.

Round 2

Reviewer 1 Report

Dear Authors,

adding about 100 lines of text to an existing article did not change my opinion. The article is interesting, but its main scientific aspect is not the subject of the Energies journal. The supplementation is not related to the applied models of renewable sources, the method of determining the energy generated by prosumers, determining the energy demand, etc. The task being implemented concerns the optimization of wireless sensor networks in smart grid. I stand by my opinion that the article should be published in a journal on a different subject. Publishing this article in Energies requires very large changes and a comprehensive expansion of the results towards the subject of the journal. 

 

Author Response

Comment:

Adding about 100 lines of text to an existing article did not change my opinion. The article is interesting, but its main scientific aspect is not the subject of the Energies journal. The supplementation is not related to the applied models of renewable sources, the method of determining the energy generated by prosumers, determining the energy demand, etc. The task being implemented concerns the optimization of wireless sensor networks in the smart grid. I stand by my opinion that the article should be published in a journal on a different subject. Publishing this article in Energies requires very large changes and a comprehensive expansion of the results towards the subject of the journal. 

Response:

Thank you so much for the valuable comment. The scope of this article is not only limited to the optimization of wireless sensor networks (WSNs) in the smart grid but also focuses on the implementation of blockchain and advanced metering infrastructure in enabling decentralized energy trading among prosumers and consumers in a smart community. Since WSN is an essential part of the microgrid, its optimization is necessary as it poses various challenges like limited power supply, computational capabilities, and network lifetime. Apart from the WSN optimization, this paper also focuses on secured and decentralized energy sharing in a smart grid system which eliminates the involvement of intermediaries and third parties, as in the case of centralized systems, and reduces the operational and transactional costs. The proposed systems are also capable of tracking the energy usage in real-time and managing the energy sharing among the households registered on a blockchain network. This not only promotes renewable energy adoption among the households but also optimizes their energy resources efficiently.

This article has been submitted to the “Energies” journal special issue on “Emerging and Advanced Green Energy Technologies for Sustainable and Resilient Future Grid”. The scope of this article fits exactly to this special issue. Thank you so much for the good suggestions and thorough reviewing of the paper.

Reviewer 2 Report

Thanks authors for their modifications. Now, the paper can be published.

Author Response

Thank you very much for your kind review.

Reviewer 3 Report

Figure 5, the authors did not answer what is the energy sharing community rate?

What is the rate of this energy forecast in this study? we do not see any technical explanation on this subject except a title.

The references 47 and 48 were not cited in the paper.

Author Response

Comment 1: Figure 5, the authors did not answer what is the energy sharing community rate?

Response: Thank you so much for the valuable comment. Figure 5 illustrates a blockchain-based advanced metering infrastructure in which the energy information is transmitted by the sensor nodes to a blockchain ledger to track the electricity usage of households and manage the P2P energy trading system among prosumers and consumers. The blockchain-based AMI is an essential component of the microgrid in the smart energy community. Within this community, the energy sharing among the households is regulated by a smart contract that takes into account the energy prices agreed among all the prosumers and manages the decentralized energy trading. The surplus energy produced by a prosumer can be shared with other households whose energy demands are not met by their renewable energy production or it can be exported to the main grid for receiving feed-in tariff and for future consumption. In the proposed web application, the rate of energy sharing is determined by the amount of energy produced by PV prosumers on a daily basis, considering the availability of energy sources and the energy consumption flexibility of prosumers. In our energy model, it is assumed that the average energy produced by a single PV prosumer in a day is 38 kWh. As per the survey by US Energy Information Administration 2019, the cost of 1 kWh of energy is 0.14 USD for households and their average power consumption in a day is 29 kWh. Based on this information and our assumption, the energy prices and supply are done on this web application. This has been explained clearly in the revised manuscript.   

Comment 2: What is the rate of this energy forecast in this study? We do not see any technical explanation on this subject except a title.

Response: Thank you so much for the valuable comment. The predictive analysis of the energy rates is out of the scope of this paper. However, we have added the factors on which energy pricing and the rate of energy sharing depend, such as availability of energy sources, demand, and economic cost, considering the energy consumption flexibility of prosumers. It is assumed that the average energy produced by a single PV prosumer on a daily basis is 38 kWh and the average energy consumed by a household in a day is 29kWh. Based on these assumptions, this paper provides a blockchain-based mechanism that facilitates decentralized energy trading among prosumers in a smart community with the aid of a smart contract. This has been explained clearly in the revised manuscript.

Comment 3: The references 47 and 48 were not cited in the paper.

Response: Thank you for pointing this out. Now all the references are cited in the revised manuscript. We apologize for this mistake.

Round 3

Reviewer 3 Report

The authors answered all my questions appropriately.

Congratulations.