Wireless Communications using Embedded MicroprocessorsView this Special Issue
Deep Learning Optimization of Microgrid Economic Dispatch and Wireless Power Transmission Using Blockchain
The purpose is to realize the decentralized microgrid economic dispatch, improve the information transparency and security of microgrid systems, and make the power grid move towards a clean, safe, efficient, and reliable development path. Deep learning optimization of microgrid economic dispatch and wireless power transmission based on blockchain technology are studied. First, the related theories and methods of microgrid systems, wireless power transmission, and deep learning optimization based on blockchain technology are introduced. Next, the microgrid economic dispatch is simulated and analyzed on a large scale. Finally, the comparison results between microgrid economic dispatch and common radio energy transmission technologies are analyzed. The results show that daily planning can better coordinate the state of distributed generation, energy storage system, and public connection. The operation results of the previous day correspond to the long-term operation economy of the microgrid. The total operation cost of the microgrid is 4668 yuan/day, and the remaining power is maintained between 500 and 600 kW, which helps to prevent excessive battery discharge, prolong battery life, and reduce operation cost. The simulation results show that the total power imbalance of the microgrid can reduce the output fluctuation of controllable load shedding of distributed generation. When the load characteristics are not important, the output fluctuation of controllable distributed generation can be reduced. The proposed economic dispatch model can optimize the data security, information storage, and information release of the microgrid and has a certain guiding role for the development of the national power grid and power industry.
In recent years, the rapid development and application of power electronic technology have laid a certain theoretical foundation for the realization of wireless power transmission (WPT). WPT technology is a revolution of traditional power transmission technology, which has very crucial advantages. The requirements for the economy, environmental protection, and reliability of power systems are increasingly higher with the increasing complexity of the power grid structure and function . Improving the energy efficiency of the power system, improving the energy structure, alleviating the contradiction between energy demand and energy utilization, as well as energy shortage and environmental protection, and maintaining the security, cleanness, and reliability of the power grid are the core high development path for the optimal operation of the power grid .
With the energy crisis, environmental pollution has also become increasingly serious. The technology of power generation using renewable energy develops rapidly. Microgrid provides an effective technical means for the use of renewable energy all over the world . The stochastic and intermittent characteristics of renewable energy make the economic dispatch of the microgrid very different from that of the traditional power systems. The economy of microgrid operation is a crucial factor for large-scale promotion, and it is also very important to study the economic distribution of microgrids. The current research focuses on the optimized distribution scheme of centralized organization, taking into account battery life, distributed generation power characteristics, microgrid, electric vehicle distribution, and so on [4, 5]. Microgrid power dispatching has also become a hot topic in the research field. Li et al. studied a parallel bidirectional power converter, which plays an important role in realizing mutual support between the two networks and improving power quality . An adaptive coordinated optimal control method of parallel Basic Physical Channel (BPC) was proposed. First, aimed at minimizing the total power loss of BPC, the economic optimal distribution scheme of power transmission between parallel BPC was calculated. Second, the primary voltage regulation controller was designed. Its outer loop could distribute dynamic transmission power between BPCs according to the power margin of each BPC to avoid BPC overload. The inner loop voltage stabilizing controller of primary voltage regulation could realize the decoupling of output variables and disturbance variables and improve the quality of dynamic voltage. Third, the secondary voltage regulation controller was designed, which could quickly restore the DC bus voltage to the rated value after the power disturbance and ensure the economic and optimal distribution of transmission power between BPC after the system reached the steady state. Finally, the stability of the proposed control method was explained, and the conclusion was verified. Wei and Chen discussed the key factors affecting the development of microgrids from the perspective of application and put forward some new suggestions for promoting the development of microgrid projects through the combination of review and promotion research . Yao et al. studied the AC/DC hybrid microgrid, taking into account the access requirements of AC/DC power supply and load, and optimized the structure of traditional distribution network . Power electronic transformer was regarded as the core of its energy management. The accurate control of voltage, current, and power flow by electrical isolation and control system made the microgrid realize a more flexible and stabler transmission mode. The power electronic transformer combines power electronic devices and high-frequency transformer, so its frequent switching leads to long time-consuming electromagnetic transient simulation. Therefore, a simplified model of dynamic response of the microgrid system under power flow and fault was proposed by simplifying the control loop and converter. The equivalent simplification method of the mathematical model was adopted. This method is simple and efficient, does not depend on the performance of the computer, and does not change the program algorithm of the software. Micallef integrated distributed generation in the form of renewable energy into a single-phase low-voltage microgrid to produce energy closer to consumers . The formation of a low-voltage microgrid can achieve high energy efficiency and improve the reliability of power supply. In these studies, microgrid operation economy is the key factor for large-scale promotion, and it is also quite essential to study the economic layout of microgrids. Although some achievements have been made, there are still multiple problems, such as high operation and maintenance cost, low stability of dispatching center, insufficient accuracy and real-time information, low prediction accuracy, and difficulty in realizing the ideal utilization of distributed energy. Unauthorized access to sensitive data and malicious tampering may directly threaten system security.
The research innovation is to focus on the distribution scheme of microgrids in organization optimization, considering battery life, distributed generation characteristics, microgrid, and electric vehicle. The methods of blockchain technology and deep learning optimization technology are analyzed and studied, and the microgrid economic dispatch simulation system is constructed. In addition, the advantages, disadvantages, and applications of the three technologies used in WPT are analyzed and evaluated. Finally, the research results are analyzed. This exploration has certain guiding significance for the safe, reliable, and economic operation of microgrids.
2. Materials and Methods
2.1. Blockchain Technology
Blockchain technology is a distributed database system jointly participated and maintained by multiple independent nodes. Each node in the blockchain system is based on a distributed and highly redundant data storage structure to store a complete blockchain. Even if a few nodes are attacked, the system can still run stably .
Figure 1 is the basic structure of a blockchain in general. It contains all the information of all nodes registered in the block body within a period of time. The information usually includes status information, transaction information, and code . It is hashed in the form of a Merkle tree and stored on the blockchain, which helps to quickly summarize and verify the accuracy and integrity of the data in the block. The header of the data block has a time stamp to indicate the time when the data is recorded. The blocks are connected according to the creation time order to form a chain structure. The header of each block contains the index of the previous block. The index hash value is conducive to tracking the source of data, increasing the difficulty of data modulation, and ensuring the reliability of data .
The decentralization of blockchain naturally adapts to the power and load balance of the microgrid. Besides, blockchain’s information openness and transparency, security and reliability, smart contract, and many other characteristics can now be applied to microgrid chain technology . The penetration of renewable energy in microgrid is increasing with the development of distributed generation technology, and the information interaction between distributed generation and local institutions becomes more frequent and adaptive. Similarly, the mainstream blockchain provides a solid foundation for data interaction of distributed generation. The information of the blockchain system is very transparent and open, and new blocks will be formed in the system every cycle. The new block contains the latest status information about the current microgrid, allowing the blockchain system to operate the power management system of the microgrid and obtain accurate historical information, comprehensive market analysis, and the most advanced process information. It improves the accuracy of renewable energy and load forecasting and ensures the economy and rationality of microgrid energy distribution. Besides, in the actual operation of a microgrid, the smart contract can ensure the automatic and safe execution of transactions on each node. Various encryption technologies are built into the blockchain to ensure data security and tamper-proof user privacy. This feature can effectively ensure the information security of distributed agents in the microgrid .
2.2. Microgrid System Based on Blockchain
A blockchain-based microgrid system mainly includes a blockchain-based energy exchange system, power management planning system, power generation, and storage system . The energy trading system based on the blockchain provides a reliable and transparent business processing logic framework and solution system for energy trading participants. The user roles related to power trading are divided into power generation (new power plants include solar power and wind power), microgrid, power consumption load side (power purchase), energy conservation, and microgrid power trading platform based on blockchain. Microgrid trading systems based on blockchains, such as power grid switching platforms, large power grids, distribution institutions and banks, and distributed energy, energy storage systems, energy converters, related power loads, and monitoring and protection equipment use a set of blockchain technology recording digital information and intelligent transactions of all parties and an innovative energy transmission and distribution system. The microgrid trading system based on blockchain is an autonomous system, which can realize self-control, protection, management, and trading . Microgrid energy distribution means optimizing distributed generation characteristics, power quality requirements, and demand management. As part of a specific control strategy, it works economically as part of the microarray and determines the optimal processing distribution and configuration of each micro power supply to achieve the microgrid . Figure 2 is a power distribution system diagram based on blockchain technology.
Based on the traditional microgrid economic division, the Evidence-Based Nursing (EBN) energy blockchain network will be integrated to make it become the information exchange and data storage center of the whole microgrid and conduct data storage, information security, and data interoperability with the blockchain for effective integration . The advantages of microgrid have been introduced into its economic transportation. The status information of distributed generation and power consumption period is monitored in real time by an intelligent electricity meter and uploaded to the EBN network. The economic dispatch scheme of the microgrid is formed by smart contract and finally realizes the stable power supply of generators in the energy consumption unit after being confirmed by the energy management system . According to the characteristics of the EBN network, a microgrid economic dispatch model based on EBN is designed, as shown in Figure 3.
Figure 4 displays the business process of microgrid economic dispatch on the energy blockchain. The specific steps are as follows. Step 1: each generator and each power user can access EBN historical data and current status information to predict their own operating status. Step 2: the node publishes its own prediction information, accepts all prediction information from other nodes, authenticates the whole network, and then saves all data. Step 3: each node invokes the smart contract based on all certified prediction information, so as to calculate the economic plan, form a planned scheduling scheme, and propagate the planned scheduling scheme to the peer-to-peer (P2P) network, waiting for other nodes to confirm. Step 4: the programming plan is saved in the smart contract in EBN after it passes the verification. If the verification fails, it is essential to return to step 3 and rerun the economic transportation calculation. Step 5: when the predefined trigger conditions are met, each power generation unit and consumer unit will automatically activate according to the schedule stored in the smart contract to end this dispatching cycle .
This exploration takes the typical microgrid structure as an example for economic allocation analysis . A microgrid is radial, including the photovoltaic system, fuel cell, wind turbine, diesel generator, micro gas turbine, battery energy storage system, large power grid, and public connection point. Among them, fuel cells, gas micro turbines, and diesel generators are controllable loads participating in the economic dispatch process. Photovoltaic systems and wind turbines are uncontrollable loads and must provide expected output information before shipment . Figure 5 is a typical microgrid structure diagram.
Figure 5 is a typical microgrid structure. It generally includes a centralized control center, distributed generation, intelligent users, energy storage equipment, and power network with self-healing (fault reconstruction) capability. With the development of microgrid research, the microgrid has gradually evolved into a group of distributed generation cluster partition networks, which can improve the original network through appropriate management and control and coordinate the operation of these distributed generators. Therefore, the definition of microgrids suitable for China should be a small modular and decentralized functional network formed by combining end-user power quality management and energy cascade utilization technology, which is based on distributed generation technology, mainly small power stations close to decentralized resources or users. The two-tier optimization strategy is adopted to achieve the economic goal of real-time deployment of microgrids on two time scales. In the daily planning stage, the node downloads the operation and maintenance cost of each distributed generation, the production of renewable energy, and load forecasting results. Besides, the node calls the smart contract of the advanced daily nonlinear economic optimization planning tool, forms the daily plan, and saves in the EBN . In the daytime programming stage, the node uploads the ultra-short-term daytime forecast of renewable energy generation load and the marginal cost of each distributed energy in combination with the daytime programming plan. The difference between the daytime plans is used as the optimization variable, and the smart contract of the daytime plan is called. The final plan is set in the form of a smart contract, saved in EBN, and scheduled automatically. Daily planning and smart contracts run every 24 hours. To create a daily scheduling plan before 00:00 every day, the trigger condition of the daily plan and smart contract is set to 23:00 every day, and the time of the daily plan is calculated. In this section, the daily plan is divided into three different stages in chronological order: morning, noon, and evening. The prediction stage involves the information disclosure of allocable units and the creation of scheduling plans .
WPT describes the process of power transmission without direct contact with wires. WPT technology saves cable cost and resistance loss for wired transmission. It has obvious advantages of convenient and flexible power supply, wide application range, and high security. Figure 6 presents the main classification methods of current WPT technology.
2.3.1. WPT Based on Electromagnetic Induction
WPT based on electromagnetic induction is similar to sensors. Both of them apply the law of electromagnetic induction to two noncontact energy transmissions. The difference is that the transformer is closely connected with the primary and secondary windings, while the WPT based on electromagnetic induction is separated from each other. Its general principle is to obtain stable DC input low-frequency AC through filter shaping and obtain high-frequency AC through the high-frequency inverter. High-frequency alternating current generates a high-frequency magnetic field. According to Faraday’s law of electromagnetic induction, the magnetic field is coupled and transmitted to the secondary side of the isolation transformer to form high-frequency AC, which is finally rectified and filtered to provide energy . Figure 7 displays the specific structure.
WPT technology based on electromagnetic induction has a short transmission distance. The transmission efficiency decreases with the increase of transmission distance, and magnetic leakage may occur. The transmission efficiency of the system depends on the coupling coefficient and quality factor of the primary and secondary measuring coils. Equation (1) is the transmission efficiency of the system: where represents the transmission efficiency of the system, is the power loss of the system, and is the output power of the system.
Equation (2) is the calculation equation of minimum value of : where represents the electric quantity of the system and represents a random constant.
At present, these types of WPT technologies are widely used to charge small portable electronic devices. For example, a typical application is the wireless charging of mobile phones, which has the advantages of simple structure, reliable technology, and low cost, as well as the disadvantages of low transmission power and short transmission distance. However, this technology is relatively mature.
2.3.2. WPT Based on Magnetic Coupling Resonance
Electromagnetic induction WPT uses air as the transmission medium, magnetic coupling resonance WPT uses a nonradiating magnetic field as the transmission medium, and the oscillator is mainly used for WPT. The resonant frequency of the device is fixed and unified, and the basic working principle is as follows. The oscillator generates a high-frequency sine wave signal, and the signal frequency matches the frequency of the resonant coil. The high-frequency oscillation signal is amplified by the power amplifier and injected into the L capacitance (LC) resonance of the transmitter. The coil then generates a magnetic field without radiating around it. At this time, the energy is transmitted to the coil at the receiving end through the combination of a nonradiating high-frequency magnetic field, forming and filtering to provide electric energy for the load. Attention should be paid to the oscillation current generated when the natural frequency of the receiving end and the frequency of the received electromagnetic wave reach the peak .
2.3.3. Microwave WPT
WPT technology is based on the microwave method and uses electromagnetic radiation to transfer energy. This is completely different from the two transmission modes mentioned above. The transmission system first converts electrical energy into a microwave of a specific frequency. A typical microwave frequency is 2.4 GHz (at present, there is no standard fixed value for microwave frequency, which needs to be determined according to specific conditions). Microwaves are transmitted and received by antennas. The microwave signal is received, and microwave conversion is performed. The rectifier can convert electric energy for wireless transmission . Figure 8 is a structure diagram of WPT based on the microwave method.
Using microwave for WPT can realize long-distance transmission, improve frequency, and realize high-energy transmission. This is the incomparable advantage of other technologies, but this technology has the disadvantages of low efficiency and directional transmission loss.
2.4. Deep Learning Optimization
Deep learning is a model of machine learning, which is a multilevel network structure that can provide complex functions. The purpose of the deep learning model is to simulate the multilevel abstract learning process of the human brain. Therefore, the structure of the multilayer neural network is proposed. The “depth” of deep learning is related to the “shallow” of shallow learning. The deep or shallow layers described here show the hierarchy of the model. Deep learning refers to the model structure of two or more hidden layers. Deep learning and shallow learning show the model structure with or without a single hidden layer. Compared with the shallow model, the deep training model can better show the characteristics of data and adapt to various complex nonlinear functions .
The deep learning model is mainly composed of three layers, including input layer, multiple hidden layers, and output layer. Each layer of the network structure contains multiple parameters. The deep learning model contains massive parameters, and many optimization algorithms have been proposed to change these parameters. The value of loss function is used to measure whether a deep learning model meets people’s requirements. The value of the loss function that meets people’s needs means that the deep learning model meets people’s needs. Otherwise, the mode will be reversed. The radio wave algorithm is executed when the value of the loss function is less than the specified accuracy. The loss functions of some common regularization terms are added to improve the relative generalization ability of the model, as shown in
represents the difference between the output value and the real value. It is the standard loss function, the regular term is , is the coefficient of the regular term, and is employed to indicate the weight of the regular term in the loss function . The larger the is, the greater the weight of the regular term is. indicates that no positive term value is added to the loss function. Regular items include the following common forms, and the following only consider the changes of weight parameters .
Equation (4) is the regular term of parameter:
The above equation shows that the minimum value point of the positive term value of the parameter is close to the origin. If the positive term value is taken to the minimum value of 0, the loss function value is close to the standard loss function value .
Equation (5) is the regular term of parameter:
The sum of the absolute values of the components of each parameter vector adopts the regular term of parameter .
Generally, the common loss function includes the mean square loss function, as shown in
In (6), is the real value and is the actual output value of the model.
Equation (7) is the cross-entropy loss function:
In (7), the number of samples is , is the real value of the th sample, and is the input value of the th sample.
Equation (8) is an exponential loss function:
The meaning of is the same as that in equation (7).
New network models and algorithms are constantly proposed with the rapid development of deep learning. In deep learning technology, Softmax function is often used as the activation function of multiclassification problems. The scalar of neuron output is mapped to the probability distribution to solve the classification problem. It solves the problem and optimizes the deep neural network . Equation (9) is the method for calculating Softmax, among which is a batch:
If the input of Softmax is , the output is , and the loss function is , (10) and (11) are the back propagation equations:
In (11), is known and needs to be calculated:
In order to maximize the probability of correct classification, it is essential to take the logarithm and minimize the cost function. Hence, if the negative number is taken, the loss function of the cost function should be
In (13), is the label and is the batch size.
From equations (9) and (13), it can be deduced that
The loss function differentiates each input :
Softmax function is a loss function in the convolutional neural network. Softmax loss has a great classification effect on multiple classification tasks, but it cannot classify if there are special categories, especially if the specific category is in two categories. In the middle of a category, it is impossible to effectively determine which category it belongs to. Center loss function can calculate multiple feature centers for each batch in the training process, and their loss functions can be calculated at the same time. The loss function can be obtained by
In (16), is the batch, is the characteristic value of the th photo, and is the center of the category to which the th photo belongs. Equation (17) is the gradient of relative to . represents the condition function. Equation (18) is the update gradient of .
Ideally, should be updated with functional changes, but it is impractical to consider all train sets in the training process. Therefore, in the training process, the median is updated with the batch as the minimum unit, and the hyperparameter is introduced to control the learning rate of the center. Equation (19) presents the update method:
3. Microgrid Economic Dispatch and WPT Analysis Results
3.1. Analysis Results of Microgrid Economic Dispatch
With a typical microgrid structure as an example, a simulation system is constructed for economic dispatch, research, and analysis. Figures 9 and 10 are diagrams of the analysis results.
Each node reads the past data and current status information of EBN; carries out weather forecast, capacity information, market situation, and overall analysis; and finally obtains the prediction power with constant hierarchy, as shown in Figure 9(c). The daily dispatching smart contract is called, and the daily dispatching plan of the microgrid is set up, as shown in Figure 9(e). According to this scheme, the total operation cost of the microgrid system is 4266 yuan/day.
The daily distribution in Figure 9(e) reveals that the period from 00:00 to 06:00 is the valley price period. The microgrid purchases as much electric energy as possible in the main network to meet the charging demand of the microgrid, and the excess energy is charged from the battery. The peak hours of electricity charges are from 6:00 a.m. to 12:00 p.m. and from 6:00 p.m. to midnight. The internal distributed microgrid power supply is used to fully power the load. Even if there is not enough internal power capacity, it is still necessary to buy power on the large power grid. Meanwhile, the battery discharges. From 12:00 to 6:00, the distributed energy and large power grid share a unified electricity charge. This is because the fuel cell power generation cost of the distributed power generation of the simulation system can be controlled to be the lowest, and the diesel generator power generation cost can be controlled to be the highest. The fuel cell is always in operation during the peak period and packaging period and runs at maximum power most of the time. The micro turbine and gas are in hot standby. The generator is in cold standby. Figure 9(f) shows the energy state of the battery. The remaining power is maintained between 500 and 600, which helps to prevent excessive battery discharge, prolong battery life, and reduce operating costs.
Figure 9(c) shows the ultra-short-term predicted power curve in a day. According to the above preday dispatching process, the daytime plan is used to call the daytime plan smart contract to get the daytime plan shown in Figure 10. With the additional cost calculation of the microgrid, it is 402 yuan/day. The main reason for the extra charge is that the daily plan is not accurate enough because it deviates from the daily forecast. Therefore, the weekly plan modifies the cost of the plan according to the more accurate ultra-short-term forecast results. Therefore, the total operation cost of a microgrid is 4668 yuan/day.
Compared with the daytime distribution scheme in Figure 10, the daytime distribution mode in Figure 9(e) has the same distribution trend in each period, which is consistent with the economics of microgrid distribution. Load response is added, especially in the daytime transportation stage. When the output is insufficient, it is essential to stop supplying power to unnecessary loads, which can improve the stability and output distribution of controllable distributed generation. It must be emphasized that the life of the power supply can be extended. Figure 10(b) shows the energy state of the battery. The remaining power is maintained between 600 and 600 to prevent serious discharge.
In order to evaluate the impact of sampling period on the reliability evaluation performance of the microgrid, different sequential Latin hypercube sampling periods (SLHS periods) are selected to calculate the reliability index of the microgrid. This method is compared with the evaluation method using random sampling. Figure 11 presents the comparison results.
Figure 11 shows the calculation results of the microgrid reliability index. It suggests that the calculation results obtained by the method proposed and the method based on random sampling are very close, and the difference of reliability index is less than 1%. When the SLHS period with different sampling periods is used for reliability calculation, the calculation results are basically the same, which proves the stability of the method. This method can significantly improve the efficiency and stability of microgrid reliability evaluation.
3.2. Comparison and Analysis of Common WPT Technologies
At present, WPT technology based on electromagnetic induction, magnetic field coupling, and microwave has obvious advantages and disadvantages. It is essential to make a comparative analysis according to the advantages and disadvantages of each position. Figure 12 is a comparison result analysis diagram.
Through technical comparison and analysis, WPT technology based on electromagnetic wave induction and WPT technology based on electromagnetic wave coupling resonance can be used in daily fields such as production and life. However, WPT technology based on the microwave method can only be useful in special fields, such as aerospace and military defense. WPT technology based on electromagnetic coupling resonance is not very mature, but it has great potential in the medium distance (more accurately, 8 times the radius of induction coil). More research is being done, such as analyzing the quantitative relationship among transmission efficiency, resonant frequency, coil size, and transmission distance and optimizing the above influence parameter values at maximum power.
The subject of economic dispatch of microgrid based on blockchain and deep learning optimization of WPT has been deeply studied. Through the analysis and research of blockchain, microgrid dispatching system based on blockchain technology, WPT, and deep learning optimization method, the microgrid economic dispatch model based on EBN is constructed. With the typical microgrid structure as an example, a simulation system for economic dispatch analysis is constructed. Meanwhile, the advantages, disadvantages, and different applications of the three technologies in the field of WPT are analyzed and compared. The reliability of the proposed model is verified. This exploration realizes the dispatching analysis and simulation of the microgrid, which provides a certain reference for related research. Restricted by the research level and other objective factors, this exploration still has some deficiencies. First, the algorithmic model of the relationship among microgrid economy, security, and reliability has not been discussed. Second, microgrid and active distribution network are new devices and lack sufficient statistical data. The collection and sorting of basic reliability data need to be further improved. In the future, the collection of operation data of these components should be strengthened, and their characteristics should be deeply analyzed, so as to establish the component model suitable for reliability evaluation.
The data used to support the findings of this study are available from the corresponding author upon request.
Conflicts of Interest
The authors declare no conflicts of interest.
Q. T. Thai, J. C. Yim, T. W. Yoo, H. K. Yoo, J. Y. Kwak, and S. M. Kim, “Hierarchical byzantine fault-tolerance protocol for permissioned blockchain systems,” The Journal of Supercomputing, vol. 75, no. 11, pp. 7337–7365, 2019.View at: Google Scholar
D. Vervoort, “Re: part 2: blockchain technology in health care,” ANZ Journal of Surgery, vol. 91, no. 4, pp. 763–763, 2021.View at: Google Scholar
Q. Xu, Z. Su, and Q. Yang, “Blockchain-based trustworthy edge caching scheme for mobile cyber-physical system,” IEEE Internet of Things Journal, vol. 7, no. 2, pp. 1098–1110, 2020.View at: Google Scholar
L. Huo, D. Jiang, S. Qi, and L. Miao, “A Blockchain-based security traffic measurement approach to software defined networking,” Mobile Networks and Applications, vol. 26, no. 2, pp. 586–596, 2021.View at: Google Scholar
M. Li, C. Lal, M. Conti, and D. Hu, “LE chain: a blockchain-based lawful evidence management scheme for digital forensics,” Future Generation Computer Systems, vol. 115, no. 6, pp. 406–420, 2021.View at: Google Scholar
P. Li, T. Guo, Y. Li et al., “An adaptive coordinated optimal control method for parallel bidirectional power converters in AC/DC hybrid microgrid,” International Journal of Electrical Power & Energy Systems, vol. 126, no. 6, p. 106596, 2021.View at: Google Scholar
P. Wei and W. Chen, “Microgrid in China: a review in the perspective of application,” Energy Procedia, vol. 158, pp. 6601–6606, 2019.View at: Google Scholar
T. Yao, Z. Li, J. Qu, Z. Li, Q. Zhao, and G. Zhao, “Research on simplified model of AC/DC hybrid microgrid for fault analysis,” Electronics, vol. 9, no. 2, p. 358, 2020.View at: Google Scholar
A. Micallef, “Review of the current challenges and methods to mitigate power quality issues in single-phase microgrids,” IET Generation, Transmission & Distribution, vol. 13, no. 11, pp. 2044–2054, 2019.View at: Google Scholar
S. Albrecht, B. Lutz, and D. Neumann, “The behavior of blockchain ventures on twitter as a determinant for funding success,” Electronic Markets, vol. 30, no. 2, pp. 241–257, 2020.View at: Google Scholar
S. Zheng, K. Liao, J. Yang, and Z. He, “Droop-based consensus control scheme for economic dispatch in islanded microgrids,” IET Generation Transmission & Distribution, vol. 14, no. 20, pp. 4529–4538, 2020.View at: Google Scholar
M. L. Richardson, “MR protocol optimization with deep learning: a proof of concept–Science Direct,” Current Problems in Diagnostic Radiology, vol. 50, no. 2, pp. 168–174, 2021.View at: Google Scholar
H. Fu, Y. Liang, X. Zhong et al., “Codon optimization with deep learning to enhance protein expression,” Scientific Reports, vol. 10, no. 1, p. 17617, 2020.View at: Google Scholar
S. Zhang, Y. Wu, C. Men, H. He, and K. Liang, “Research on open CL optimization for FPGA deep learning application,” PLoS One, vol. 14, no. 10, p. e0222984, 2019.View at: Google Scholar
Z. X. Li, W. Song, and Y. Tian, “Exchange rate forecasting and portfolio optimization based on deep learning and evolutionary computation,” Journal of Zhengzhou University (Engineering Edition), vol. 40, no. 1, pp. 92–96, 2019.View at: Google Scholar
N. A. Kallioras, G. Kazakis, and N. D. Lagaros, “Accelerated topology optimization by means of deep learning,” Structural and Multidisciplinary Optimization, vol. 62, no. 3, pp. 1185–1212, 2020.View at: Google Scholar
H. Deng and A. C. To, “Topology optimization based on deep representation learning (DRL) for compliance and stress-constrained design,” Computational Mechanics, vol. 66, no. 2, pp. 449–469, 2020.View at: Google Scholar
S. J. Berlin and M. John, “Particle swarm optimization with deep learning for human action recognition,” Multimedia Tools and Applications, vol. 79, no. 25, pp. 17349–17371, 2020.View at: Google Scholar
K. Singh and R. K. Kapania, “Accelerated optimization of curvilinearly stiffened panels using deep learning,” Thin-Walled Structures, vol. 161, no. 3, p. 107418, 2021.View at: Google Scholar
G. Agarwal and H. Om, “Performance of deer hunting optimization based deep learning algorithm for speech emotion recognition,” Multimedia Tools and Applications, vol. 80, no. 4, pp. 1–32, 2021.View at: Google Scholar
S. K. Oruganti, F. Liu, D. Paul et al., “Experimental realization of Zenneck type wave-based non-radiative, non-coupled wireless power transmission,” Scientific Reports, vol. 10, no. 1, p. 925, 2020.View at: Google Scholar
X. Wang, C. Lu, C. Wang et al., “Methods for testing the performance of long-distance wireless power transmission systems,” EURASIP Journal on Wireless Communications and Networking, vol. 2020, no. 1, 2020.View at: Publisher Site | Google Scholar
K. Takahashi, T. Yamada, and Y. Takemura, “Circuit parameters of a receiver coil using a Wiegand sensor for wireless power transmission,” Sensors, vol. 19, no. 12, p. 2710, 2019.View at: Google Scholar
D. K. Freeman and S. J. Byrnes, “Optimal frequency for wireless power transmission into the body: efficiency versus received power,” IEEE Transactions on Antennas and Propagation, vol. 67, no. 6, pp. 4073–4083, 2019.View at: Google Scholar
S. Nimura, D. Furusu, and M. Tamura, “Improvement in power transmission efficiency for cavity resonance-enabled wireless power transfer by utilizing probes with variable reactance,” IEEE Transactions on Microwave Theory and Techniques, vol. 68, no. 7, pp. 2734–2744, 2020.View at: Google Scholar
J. Hur and H. Choo, “Design of a small array antenna with an extended cavity structure for wireless power transmission,” Journal of Electromagnetic Engineering and Science, vol. 20, no. 1, pp. 9–15, 2020.View at: Google Scholar
G. J. Gwon and Y. Kwon, “Enhancement of wireless power transmission efficiency and flexibility via an optimized three-dimensional coupled magnetic resonance system with double transmitter coil,” Journal of Electrical Engineering & Technology, vol. 16, no. 3, pp. 1415–1426, 2021.View at: Google Scholar
L. Shi, N. Rasool, H. Zhu, K. Huang, and Y. Yang, “Design and experiment of a reconfigurable magnetic resonance coupling wireless power transmission system,” IEEE Microwave and Wireless Components Letters, vol. 30, no. 7, pp. 705–708, 2020.View at: Google Scholar
X. Li, K. M. Luk, and B. Duan, “Aperture illumination designs for microwave wireless power transmission with constraints on edge tapers using Bezier curves,” IEEE Transactions on Antennas and Propagation, vol. 67, no. 2, pp. 1380–1385, 2019.View at: Google Scholar