Intelligent Dispatching Method of Short-term Load in Distributed Power System Using Deep Learning

Accurate short-term electric load forecasting is significant for the smart grid. It can reduce electric power consumption and ensure the balance between power supply and demand. In this paper, the Stacked Denoising Auto-Encoder (SDAE) is adopted for short-term load forecasting using four factors: historical loads, somatosensory temperature, relative humidity, and daily average loads. The daily average loads act as the baseline in final forecasting tasks. Firstly, the Denoising Auto-Encoder (DAE) is pre-trained. In the symmetric DAE, there are three layers: the input layer, the hidden layer, and the output layer where the hidden layer is the symmetric axis. The input layer and the hidden layer construct the encoding part while the hidden layer and the output layer construct the decoding part. After that, all DAEs are stacked together for fine-tuning. In addition, in the encoding part of each DAE, the weight values and hidden layer values are combined with the original input layer values to establish an SDAE network for load forecasting. Compared with the traditional Back Propagation (BP) neural network and Auto-Encoder, the prediction error decreases from 3.66% and 6.16% to 2.88%. Therefore, the SDAE-based model performs well compared with traditional methods as a new method for short-term electric load forecasting in Chinese cities.

Optimal management of hybrid PV/fuel cell/battery power system: A comparison of optimal hybrid approaches

A hybrid Improved Quantum-behaved Particle Swarm Optimization–Simplex method (IQPSOS) to solve power system load flow problems

Short-term power load prediction is not only a critical part of power system dispatching but also an essential task for power marketing, grid planning, and other management departments; deep learning is an artificial intelligence method that has gained extensive attention in recent years; this paper selected three models with the most typical signs of deep learning recurrent neural networks and studied their model performance for short-term power load predictions, including power load data preprocessing, feature selection of prediction models, determination of model parameters, and comparison of prediction characteristics between shallow and deep networks, to explore and prove the applicability of deep learning neural networks in short-term power load predictions, ultimately improving the accuracy of short-term load predictions.KeywordsPower loadDeep learningPrediction

The Aerospace Corporation has independently developed a distributed electric power system for picosatellites. This distributed power system architecture is being implemented on a number of upcoming picosatellite missions. The first implementation of this architecture will be flown on the PowerSphere Flight Experiment and the Pico Satellite Inflatable Reflector Experiment (PSIREX). To date all of the picosatellite missions do not have attitude control or deployable solar array structures. Solar cells are body mounted on the various sides of the picosatellite. The architecture was conceived to solve the problem of obtaining the maximum amount of electric power from a solar array, that has multiple panels, that are not arranged on a single planar surface. Aerospace Corporation researchers working on developing viable power systems for picosatellites were awarded patents 6,127,621 October 2, 2000 Power Sphere and 6396167 B1 May 28, 2002 Power Distribution System for this unique solution for distributed power system architecture for a multifaceted solar array. The authors have developed a prototype of the ring bus and have completed initial testing of its performance.

A model integrating Particle Swarm Optimization (PSO) and support vector machines (SVM) is presented to forecast short-term load of electric power systems in this paper. PSO is a method for finding a solution of stochastic global optimizer based on swarm intelligence. Using the interaction of particles, PSO searches the solution space intelligently and finds out the best one. The PSO-SVM method proposed in this paper is based on the global optimization of PSO and local accurate searching of SVM. Practical example results indicate that the application of the PSO-SVM method to short term load forecasting of power systems is feasible and effective. And to prove the effectiveness of the model, other existing methods are used to compare with the result of SVM. The results show that the model is effective and highly accurate in the forecasting of short-term power load.

In order to build a low-carbon and environmentally friendly energy system, a large number of distributed generation connected to the grid, which has brought great challenges to the stable operation of the distribution network. The intermittency and volatility of distributed generation lead to repeated voltage violations in power system. At the same time, the frequent switching of transformer taps and capacitor switches greatly increases the economic cost. This paper introduces the method of deep reinforcement learning to transform the Volt-Var optimize problem into a Markov decision process with continuous action space. The soft actor-critic algorithm is used to train the agent to control the static var compensator and the smart inverter to optimize the reactive power of the distribution network. Comparative studies on IEEE 12-node and 33-node test systems demonstrate the performance of the proposed solution.

The traditional home power system is not suitable for the current complex household appliances for its low efficiency, which need a new power system. In this paper, a future home power system with contactless power transmission (CPT) technologies is proposed. Besides, with the consideration of using renewable power, a solar power hybrid generation system is integrated in the future power system. In addition, the distributed power system is introduced to the proposed power system instead of traditional centralized power method. The distributed power system is more stable and efficient. On the other hand, thanks to the CPT technologies, household appliances will be more safe and friendly by using CPT systems than before by eliminating cords and plugs. A simulation by PSIM DEMO version of the proposed future power system shows that the proposed concept is feasible. Furthermore, a CPT prototype has been fabricated to analyze the circuit characteristics of the CPT system.A novel concept of home power system is proposed, elaborated in detail and simulated by computer software. DOI: http://dx.doi.org/10.11591/telkomnika.v11i12.3692 Full Text: PDF

As one of the countermeasures to solve the increase problem of the fault current in a power system, the superconducting fault current limiter (SFCL) has been noticed as the attractive device because it has no power loss in a normal time and can limit the fault current rapidly without additional device to detect the fault current. However, the introduction of the SFCL in a power distribution system affects the protection coordination between the existing protective devices such as the over-current relay (OCR) and the recloser, which causes the operation of the protective devices to get out of their original setting values. Therefore, the analysis on the effect of the SFCL introduction on the existing protective devices is required and the method to keep the protection coordination between the protective devices in a power system with the SFCL should be first considered. In this paper, the protection coordination experiments between the OCR and the recloser in the simulated power distribution system with the SFCL were performed. Based on the experimental results, the influence of the SFCL on the operation of these protective devices was analysed and the methods to keep the protection coordination between the OCR and the recloser in a power distribution system with the SFCL were discussed.

The short-term interval load forecasting accuracy of the power system has an important impact on the power system operation. A short-term load forecasting model based on ALO-ML-ELM is proposed for short-term load-interval forecasting without considering the load trend and inaccurate interval construction. Firstly, the weather, temperature, and historical load data are processed and normalized by taking the load trend influencing factors into account. Then the load power interval is constructed by fitting a normal distribution to the historical load error data, obtaining the relevant parameters, and generating random noise. Next, the short-term load-interval prediction model is established by optimizing the input layer weights of the multi-layer extreme learning machine (ML-ELM) by Ant Colony Optimization (ALO). Finally, it is validated by the actual load data of a city in China. The simulation results show that the proposed method has higher interval coverage and accuracy.

Accurate forecasting of power system short -term load has been one of the most important issues in the electricity industry. And the forecasting accuracy is influenced by many unpredicted factors. Because of the non-linear features of short-term power load, the paper uses support vector machines (SVM) technology for the short-term electricity load forecast. The method can better solve such practical problems as small samples, nonlinearity, high dimensionality and local minimization, which can greatly enhancing its ability to handle non-linear. To solve the problems of SVM in training for large-scale convergence, such as slow convergence, greet complexity, particle swarm optimization(PSO) is proposed for the secondary planning problem to enhance SVM computing speed. The improved SVM is applied to short-term load forecasting, empirical studies show that the method has a high prediction accuracy and faster computing speed.

In this paper, a Q learning quantum-behaved particle swarm optimization (QPSO) method is proposed to solve the economic dispatch (ED) problem in power systems, whose objective is to simultaneously minimize the generation cost rate while satisfying various equality and inequality constraints. The proposed method enhance the global search ability of the algorithm. The feasibility of the QlQPSO method is demonstrated by three different power systems, compared with the GIQPSO, the MeQPSO, and the SlQPSO in terms of the solution quality, robustness and convergence property. The simulation results show that the proposed QlQPSO method is able to obtain higher quality solutions stably and efficiently in the ED problem than other tested optimization algorithm.

A Solid-State Transformer as Enabling Technology for Microgrids for Integration into the Power Distribution System: An Overview

This paper shows a method that reconfigures the distribution power system by using branch exchange algorithm. The optimal reconfiguration method calculates line loss and voltage condition in all possible power system configurations depending on line on/off, and it searches for the most appropriate optimal configuration. However, it is not easy to search for the solution for the most appropriate optimal configuration in distribution power system for a short time because there are a number of connection & sectionalized switches in distribution power system and it should comply with the condition of radial operation. Therefore, this research shows the method that not only reduces the whole system loss by using branch exchange algorithm which has the advantage in time, but also complies with the condition of radial operation and the constraint condition of voltage.

This research explores the integrated management of water distribution systems (WDS) and power distribution systems (PDS) to improve their resilience to extreme scenarios. This study delves into the dynamics of a locally managed PDS as an example of extreme operational conditions. The primary objective is to minimize load shedding (LS) in the PDS through strategic load shifting in the interconnected WDS, demonstrating the potential of cooperative decision making between the two critical systems. The optimization framework offers a novel approach to managing flexible resources during emergencies by utilizing the mutual links between a WDS and a PDS. Typically, WDSs and PDSs are operated by different operators such that cooperation is limited. This study presents how communication based on limited information sharing between the two systems is sufficient to increase resilience and improve the systems’ functionality, emphasizing the advantages of cooperative decision making. This paper highlights the significance of cross-sectoral collaboration, presenting a viable pathway for managing local infrastructure systems under extreme conditions while ensuring uninterrupted service delivery to communities.