Fast and accurate load parameter identification has a large impact on power systems operation and stability analysis. This article proposes a novel Imitation and Transfer Q-learning (ITQ)-based method to identify parameters of composite constant impedance-current-power (ZIP) and induction motor (IM) load models. Firstly, an imitation learning process is introduced to improve the exploitation and exploration processes. Then, a transfer learning method is employed to overcome the challenge of time-consuming optimization when dealing with new identification tasks. An associative memory is designed to realize dimension reduction, knowledge learning and transfer between different identification tasks. Agents can exploit the optimal knowledge from source tasks to accelerate the search rate in new tasks and improve solution accuracy. A greedy action selection rule is adopted for agents to balance the global and local search. The performance of the proposed ITQ approach has been validated on a 68-bus test system. Simulation results in multi-test cases verify that the proposed method is robust and can estimate load parameters accurately. Comparisons with other methods show that the proposed method has superior convergence rate and stability.
The most widely used high power industrial lasers are Nd:YAG and carbon-dioxide lasers. Chemical oxygen iodine laser (COIL), whose wavelength (1.315 micrometer) is between that of YAG (1.06 micrometer) and carbon-dioxide (10.6 micrometer) lasers, is another high power laser for industrial applications. The cutting capability of these lasers is investigated. The cut depth depends strongly on the absorptivity of materials, kerf width and cutting speed. Absorptivity is an unknown parameter for which experimental data at high temperatures are unavailable. Theoretical values of the absorptivities of various metals are obtained using Hagen-Ruben's relation. It is found that the absorptivity of metals is linearly proportional to the square root of resistivity and inversely proportional to the square root of the wavelength. The absorptivities of COIL ad YAG lasers are 2.84 and 3.16 times larger than that of carbon-dioxide laser, respectively. Based on the theoretical values of absorptivity, the cut depth of metals are analyzed for various laser powers, cutting speeds for these lasers. Due to the wavelength dependence of absorptivity, the cut depths for COIL and YAG lasers are expected to be 2.84 and 3.16 times deeper than that for carbon-dioxide laser.
Advancements in information and communication technologies have revolutionized monitoring and control capabilities within smart grids. However, it also brings new vulnerabilities to data acquisition systems and state estimation functions, which attackers can subtly tamper with the measurement data through compromising the communication network. Moreover, the high penetration of renewable energy sources with the inherited characteristics of uncertainty and variability further complicates the design of effective intrusion detection systems. In this paper, a Bayesian deep learning-based approach is developed to detect cyber attacks and maintain the security of smart grids. Our method specifically addresses the prevalent issue of imbalanced data in real power systems, which arises from the predominance of normal system operations over compromised or attacked states. Employing a novel Bayesian GAN-based technique, our approach successfully discriminates between secure and compromised measurement data, even in scenarios with significant data imbalance. Furthermore, the proposed method accommodates various practical application factors, ensuring accurate intrusion detection despite the presence of measurement noise. The feasibility and effectiveness of the proposed detection mechanism are validated by testing on IEEE 13-node and 123-node test systems. Simulation results and comparisons with literature methods demonstrate the superiority of proposed cybersecurity solutions.
An accurate awareness of characteristics of the electromechanical oscillation modes is essential for the secure operation of the power systems. The random decrement technique (RDT) is used with the multi reference point complex index (PRCE) which was formerly applied to analyze the vibration characteristics of mechanism structures to estimating the inter-area electromechanical modes during ambient condition, including the frequencies, damping ratios and the mode shapes. A Monte Carlo simulations were conducted to evaluate the performances of the RDT-PRCE method in a statistical way. Simulation results based on the 16-machine system indicate that the proposed method can provide a more accurate estimated results compared with the RDT-TLS-ESPRIT method and NExT-ERA method, also, the RDT-PRCE method is highly robust against measurement noise. The RDT-PRCE method was also applied to actual measurements from Sichuan power grid of China. The results show that the RTD-PRCE method is valid and fast enough to be implemented in the real-time use.
This paper describes two data driven methods to extract the ring-down signatures of the random responses for the study of low frequency electromechanical oscillations. The two methods are: random decrement technique (RDT) and the natural excitation technique (NExT). The objective is to compare and contrast the performance of the methods. The correlation coefficients between the extracted ring-down signal and the impulse signal, the accuracy of damping ratio and frequency estimated by RDT-ERA method and the NExT-ERA method are the measures of the performance of the algorithms. Their performances are evaluated using an ideal second-order system, and PMU measurements from Sichuan power grid of China.
With the rapid development of social economy and the Internet, the network education is becoming a way of teaching which has a wide application range and covering larger area.Virtual learning community (VLC) is a combination of computer technology, psychology, pedagogy and other multi-disciplinary research field and actually a new model of network education.However, the teaching data of VLC are often disorderly, fragmentary, mixed and its value is also not easy to detect.The using of data mining technology will solve this kind of problems and bring many unexpected benefits support the teaching of the VLC.This paper reports on the analysis of learning behavior of the VLC and how to extract the feature vector of learning.The fuzzy c-means clustering algorithm is applied to analyze the learning behavior and divide the students of the VLC by the feature of them.Then some targeted teaching guidance can be made for each group.This kind of grouping strategy is to be found feasible and achieved good effect by simulation experiment.
The frequency stability of power system is usually jeopardized by the active power deficit for major disturbances. To deal with these events, adaptive load shedding schemes are effective, and the steady frequency prediction is considered as the prerequisite. However, none of previous predictive methods has taken into consideration the nonlinear segment that may distort system frequency response. This paper investigates the effect of speed governor deadband and presents a method to evaluate its effect on the steady frequency. Based on the wide area measurement system measurements, a predictive algorithm of steady frequency is studied. It utilizes the modified regulation coefficient to consider the governor deadband effect and employs instantaneous measurements of postdisturbance to calculate the steady frequency directly. Simulations on the IEEE-9 system and a practical system of China have solidified the effectiveness of the proposed algorithm.
In this paper, the chaos control of Lozi mapping is carried out through the improved OGY method. By establishing linear mapping and applying the pole placement technique of the linear control theory, small time-dependent perturbations of a control parameter are selected. Unstable periodic orbits embedding in the chaos attractor are examined firstly, and then the period one orbit is chosen as a control target. When map point wanders to the neighborhood of the periodic point, system control parameter is perturbed. The unstable period one point is controlled to be the stable periodic orbit. At the same time, the different choice of the regulator poles is analyzed. Using numerical simulation, the effectiveness of the method is demonstrated.
Emergency frequency control is one of the most critical approaches to maintain power system stability after major disturbances. With the increasing number of grid-connected renewable energy sources, existing model-based methods of frequency control are facing up with challenges of computational speed and scalability for large-scale systems. In this paper, the emergency frequency control problem is formulated as a Markov Decision Process and solved through a novel distributional deep reinforcement learning (DRL) method, namely the distributional soft actor critic (DSAC) method. Compared with other reinforcement learning methods that only estimate the mean value, the proposed DSAC model estimates the distribution of value function over returns. This advancement can lead to more insights and knowledge for the agent, with the benefit of a much faster and more stable learning process, and the improved frequency control performance. The simulation results on IEEE 39-bus and IEEE 118-bus systems demonstrate the effectiveness and robustness of proposed models, as well as the advantage compared to other state-of-the-art DRL algorithms.
When the single-phase to ground fault occurs, the initial current traveling wave of fault line and normal lines will change steeply, if the capacitor of bus isn't considered. When the capacitor of bus switches, the initial current traveling wave of normal line will change slightly, but that has no effect on the fault line which still keeps abruptly change. Based on that, this paper proposes a new line selection method to identify the fault line by using the capacitor switching strategy when the fault occurs. Though complex transform, many disadvantages of real wavelet transform can be avoided. This method can enlarge the differences between normal line and fault line, and reinforce the detection accuracy in a great extent. The simulation tests verify the efficiency and sensitivity of the proposed algorithm.
'/%3e%3cuse xlink:href='%23a' transform='rotate(23.036 .093 25.536)'/%3e%3cuse xlink:href='%23a' transform='rotate(45.87 1.273 16.18)'/%3e%3cuse xlink:href='%23a' transform='rotate(69.945 .996 12.078)'/%3e%3cuse xlink:href='%23a' transform='rotate(20.66 -19.689 31.932)'/%3e%3c/svg%3e)
