Search

Previously works on analysing imperfect electricity markets have employed conventional game-theoretic approaches. However, such approaches necessitate that each strategic market player has full knowledge of the operating parameters and the strategies of its rivals as well as the computational algorithm of the market clearing process. This unrealistic assumption, along with the modeling and computational complexities, renders such approaches less applicable for conducting practical multi-period and multi-spatial equilibrium analysis. This paper proposes a novel multi-agent deep reinforcement learning (MA-DRL) based methodology, combining multi-agent intelligence, the deep policy gradient (DPG) method, and an innovative long short term memory (LSTM) based representation network for optimizing the offering strategies of multiple self-interested generation companies (GENCOs) as well as exploring the market outcome stemming from their interactions. The proposed approach is tailored to align with the nature of the examined problem by posing it, for the first time, in multi-dimensional continuous state and action spaces, enabling GENCOs to receive accurate feedback regarding the impact of their offering strategies on the market clearing outcome, and devise more profitable bidding decisions by exploiting the entire action domain, and thereby facilitates more accurate equilibrium analysis. The proposed LSTM-based representation network extracts discriminative features which further improves the learning performance and thus promises more profitable offerings strategies for each GENCO. Case studies demonstrate that the proposed method i) achieves a significantly higher profit than state-of-the-art RL methods for a single GENCO's optimal offering strategy problem and ii) outperforms the state-of-the-art equilibrium programming models in efficiently identifying an imperfect market equilibrium with/without network congestion. Quantitative economic analysis is carried out on the obtained equilibrium.

Bi-level optimization and reinforcement learning (RL) constitute the state-of-the-art frameworks for modeling strategic bidding decisions in deregulated electricity markets. However, the former neglects the market participants’ physical non-convex operating characteristics, while conventional RL methods require discretization of state and / or action spaces and thus suffer from the curse of dimensionality. This paper proposes a novel deep reinforcement learning (DRL) based methodology, combining a deep deterministic policy gradient (DDPG) method with a prioritized experience replay (PER) strategy. This approach sets up the problem in multidimensional continuous state and action spaces, enabling market participants to receive accurate feedback regarding the impact of their bidding decisions on the market clearing outcome, and devise more profitable bidding decisions by exploiting the entire action domain, also accounting for the effect of non-convex operating characteristics. Case studies demonstrate that the proposed methodology achieves a significantly higher profit than the alternative state-of-the-art methods, and exhibits a more favorable computational performance than benchmark RL methods due to the employment of the PER strategy.

In traditional Chinese medicine processing domain, the ‘firepower’ and ‘proper roasting’ are vital for medicinal efficacy of Chinese herbs. We take phellodendron microscopic images to determine if the roasting is appropriate according to some features of crystal fiber and stone cell, these features include color, geometrical appearance, and so on. The first step of feature analysis is to segment objects accurately. In the paper, we propose an efficient unsupervised algorithm to segment stone cell and crystal fiber in phellodendron microscopic images automatically without any human's intervene. Firstly, the superpixels is adopted, then the method extracts features such as RGB, Sobel, Harris and so on, clustering superpixels patches according to the characteristic of phellodendron, lastly morphological operations are implemented to produce segmented stone cell and crystal fiber. The experimental evaluation shows that the proposed algorithm can segment objects automatically with about 87% precision and 93% recall.

Rotating emergency generators (EG) have been increasingly installed to provide backup power in cases of blackouts. However, conventional rotating generators cannot provide fast pickup of load as soon as a blackout is detected. The paper presents a hybrid UPS system based on a battery energy storage system (BESS) and a rotating EG that can provide Class 2 response to the load during the power transfers between the load and the utility as per the Japanese standard on UPS JEC2433-2016. The same idea is also shown to work for cost savings by switching to backup power for areas with multi-part or time-of-use tariffs. The design process including controller hardware in the loop (CHIL) testing with real time simulators is also presented as it makes the design process more efficient. The process was used to test compliance with JEC2433-2016 in addition to speeding up the controller validation and testing step in the product design and development cycle. The proposed hybrid UPS is validated using CHIL and hardware testing.

This paper shows the development of a power supply system to solve the anomalous voltage sag, power failure and peak cut demand. This power supply system includes an UPS (Uninterruptible Power Supply), a standby generator and a Hybrid Switch, which is the integration of mechanical switches and IGBTs with good performance of low conducting loss and fast opening time. The results of a real-time simulation demonstrate its voltage compensation can perform within 2 ms response time when power failure occurs and transfer without voltage down between micro grid operation and gird-connected operation. Also, they are successfully validated by using a prototype system.

Previous work has demonstrated the significant benefits of demand flexibility for low-carbon electricity systems. However, given the deregulation of the electricity industry, this flexibility also affects the business case of retailers, an aspect which has not yet been thoroughly investigated. This paper proposes a new multi-period bi-level optimization model capturing the interactions between a retailer's offered tariffs to the served consumers, the consumers' flexibility and the wholesale market in an integrated fashion. This bi-level problem is efficiently solved after converting it to a Mathematical Program with Equilibrium Constraints (MPEC), and subsequently to a Mixed-Integer Linear Problem (MILP). Case studies demonstrate that the time-shifting flexibility of the consumers reduces the retailer's revenue, since consumers can respond more effectively to strategic retail pricing patterns, but also reduces the retailer's cost, since it leads to flatter wholesale demand and price profiles, an effect which is not captured by previous work. Finally, the results indicate that the impact of this flexibility on the retailer's business case depends on the level of competition with other retailers in the market.

To minify the scale of Petri net and improve both accuracy and adaptability of power system fault diagnosis, a model of fuzzy hierarchical Petri net with multi subnets is proposed, in which different effects of main protection, local backup protection and remote backup protection are considered, and different weights are assigned to output arcs transited to termination places in protection subnets, and then the Gaussian function to solve output probabilities of the transition is modified The reasoning process based on matrix operations is given. The effectiveness of the proposed fuzzy hierarchical Petri net is verified by results of multi groups of test cases, and test results show that correct fault diagnosis results can be obtained under incomplete information or variation of network topology.

Bi-level optimization constitutes the most common mathematical methodology for modeling the decision-making process of strategic generation companies in deregulated electricity markets. However, previous models neglect the physical non-convex operating characteristics of generation units, due to their inherent inability to capture binary decision variables in their representation of the market clearing process, rendering them problematic in the context of markets with complex biding and unit commitment clearing mechanisms. Aiming at addressing this fundamental limitation, this paper delves into deep reinforcement learning-based approaches by proposing a novel deep Q network (DQN) method and enabling explicit incorporation of these non-convexities into the bi-level optimization model. Case studies demonstrate the superior performance of the proposed method over the conventional Q-learning method in devising more profitable offering strategies.