Neutral Reactor Research Articles

Detection of secondary arc extinction and auto-reclosing in compensated AC transmission lines based on machine learning

Detecting and recognizing the secondary arc extinction is considered as a backbone of the single pole auto reclosing schemes. This paper proposes a multi-channel convolutional network to detect the secondary arc extinction during a long AC power transmission line arc fault. In contrast to the conventional methods, the proposed Machin learning-based method can automatically learn the features from the raw data given as input. Multivariate time series data of phase voltage, current and neutral reactor signals were used as the input to the proposed model. The features are extracted in the convolution layer instead of using hand-crafted feature extraction like most of the existing research. The softmax classifier is used to detect fault and secondary arc extinction. Matlab Simulink is used to simulate the test system and collect the dataset to evaluate the proposed neural network model. To verify the generalizability of the proposed architecture datasets were collected at various locations and different phase lines. The proposed algorithm was able to detect the secondary arc extinction with 98.26% accuracy and able to give signal for auto-recloser logic with the maximum window length of three cycles. The result shows that the proposed neural network architecture is accurate and robust to detect the secondary arc extinction further can be used as a signal for a single pole adaptive auto recloser scheme.

Hybrid Series/Parallel All-Nonlinear Dynamic-Static Neural Networks: Development, Training, and Application to Chemical Processes

This paper presents the development of data-driven hybrid nonlinear static-nonlinear dynamic neural network models and addresses the challenges of optimal estimation of parameters for such hybrid networks. A parallel static-dynamic neural network and two variants of series networks, specifically, nonlinear static-nonlinear dynamic and nonlinear dynamic-nonlinear static networks, are investigated in this work. Performances of the proposed fully nonlinear hybrid series and parallel network models are compared with the existing state-of-the-art data-driven models like long–short-term memory networks and gated recurrent unit models as well as DABNet-type linear-dynamic-nonlinear-static neural network. Algorithms are developed for training the series and parallel hybrid networks where the static and dynamic networks can be trained independently by different algorithms. These algorithms offer flexibility for incorporating different types of static and dynamic network architectures and their training algorithms thus offering tradeoff between computational expense and accuracy for highly nonlinear systems. In addition to the typical training objective that minimizes some squared error, an objective function is considered that penalizes overfitting and uncertainty in parameter estimates. Performances of the proposed algorithms are evaluated for three nonlinear example problems─a two-tank pH neutralization reactor, the widely used Van de Vusse reactor, and a pilot plant for postcombustion CO2 capture using monoethanolamine solvent. Computational expense and convergence performance of the proposed network architectures and their training algorithms are presented.

Determining optimal neutral reactor to limit overvoltages and duration of secondary arc of transmission lines

AbstractEven though shunt compensation of transmission lines (TLs) is a common practice to limit the overvoltages stemming from the Ferranti effect, it may cause transient and steady‐state resonant overvoltages when one phase is left open due to the single‐pole‐open condition following the temporary phase‐to‐ground (Ph‐g) fault. Furthermore, the existing shunt reactors potentially increase the duration of the secondary arc current at the fault point and hence prevent offering fast single‐pole reclosing. To exploit the mentioned issues, in recent studies, a neutral reactor is introduced, which is installed in the neutral point of the shunt reactors; nonetheless, obtaining the appropriate amount of it is remarkably challenging. In this paper, an optimal neutral reactor (ONR) is determined based on an optimization problem, which minimizes the neutral reactor cost subject to limiting both resonant overvoltages and the duration of the secondary arc current. In modelling, the Laplace domain is employed rather than the phasor domain to consider both transient and steady‐state conditions since this technique allows having better control over transient overvoltages as well as transient secondary arc.

Nonlinear model predictive control algorithm with iterative nonlinear prediction and linearization for long short-term memory network models

In this paper, a practical nonlinear model predictive control with iterative nonlinear prediction and linearization is proposed, considering a long short-term memory (LSTM) artificial neural network (PNMPCi-LSTM) as process model for making the predictions. The prediction model is divided into two portions, the base output prediction, obtained with the LSTM nonlinear model, and the incremental output prediction, obtained using a linearized version of the LSTM model. The base response and the dynamic matrix of the system, which is obtained using the linearized version, are used to find an optimal control effort by solving a quadratic programming problem. This procedure is performed iteratively by updating the base input with the candidate control effort until the incremental response term is small enough compared with the base response term. The advantages of the proposed method in terms of performance and computing times are illustrated using the control of a simulated nonlinear neutralization reactor. For the evaluated case study, the results show that by using the proposed iterative procedure the closed-loop performance measured using the integral absolute error is improved by 8% for a setpoint tracking scenario while keeping the computation times within reasonable levels. In addition, the results support the idea that the proposed PNMPCi-LSTM is an alternative to implement a nonlinear MPC with reasonable computation times.

Initialisation of Optimisation Solvers for Nonlinear Model Predictive Control: Classical vs. Hybrid Methods

In nonlinear Model Predictive Control (MPC) algorithms, the number of cost-function evaluations and the resulting calculation time depend on the initial solution to the nonlinear optimisation task. Since calculations must be performed fast on-line, the objective is to minimise these indicators. This work discusses twelve initialisation strategies for nonlinear MPC. In general, three categories of strategies are discussed: (a) five simple strategies, including constant and random guesses as well as the one based on the previous optimal solution, (b) three strategies that utilise a neural approximator and an inverse nonlinear static model of the process and (c) four hybrid original methods developed by the authors in which an auxiliary quadratic optimisation task is solved or an explicit MPC controller is used; in both approaches, linear or successively linearised on-line models can be used. Efficiency of all methods is thoroughly discussed for a neutralisation reactor benchmark process and some of them are evaluated for a robot manipulator, which is a multivariable process. Two strategies are found to be the fastest and most robust to model imperfections and disturbances acting on the process: the hybrid strategy with an auxiliary explicit MPC controller based on a successively linearised model and the method which uses the optimal solution obtained at the previous sampling instant. Concerning the hybrid strategies, since a simplified model is used in the auxiliary controller, they perform much better than the approximation-based ones with complex neural networks. It is because the auxiliary controller has a negative feedback mechanism that allows it to compensate model errors and disturbances efficiently. Thus, when the auxiliary MPC controller based on a successively linearised model is available, it may be successfully and efficiently used for the initialisation of nonlinear MPC, whereas quite sophisticated methods based on a neural approximator are very disappointing.

Neutral Reactor Structures for Improved Single Phase Auto Reclosing in Multi-Circuit Multi-Voltage Transmission Lines

In this paper, novel neutral reactor structures are proposed to mitigate the secondary arc current (SAC) in non-conventional Multi-circuit Multi-voltage Transmission lines (MCMVTLs). The parameters of the reactors in the structure are optimized using the Non-dominated Sorting Genetic Algorithm II (NSGA-II). The optimization algorithm, coded in Python, regulates the multiple runs of an Electromagnetic Transients (EMT) simulation program to minimize the secondary arc current. The optimization trials are run in parallel on a 64 core computer to minimize the solution time. A detailed arc model was tested on the EMT program (PSCAD). Using this model and optimized parameters in the proposed neutral reactor schemes, the SAC is indeed extinguished in acceptable times. The SAC and TRV for neutral reactors with optimized parameters are up-to 68 and 77%, respectively, lower than the ones with parameters calculated by the conventional approach which ignores inter-circuit coupling. Meanwhile, the extinction times are up-to 46% faster. As a result, single-phase auto-reclosing in MCMVTLs is viable.

Algorithm for controlling the process of buffer wastewater neutralization

In Ukraine, the condition of surface water near industrial enterprises is extremely critical. The operation of enterprises leads to intensive water pollution with industrial and domestic wastewater. Therefore, improving the quality of treatment facilities through the introduction of automated control systems is an urgent problem.
 For the operation of automated control systems for typical cleaning processes, a software package is required, which is developed on the basis of appropriate algorithmic software and mathematical models of processes. To obtain them, methods of mathematical and simulation modeling and block diagram method of algorithmization were used.
 In order to assess the quality of the developed algorithm for controlling the process of buffer wastewater neutralization during operation, a comparison of control system operation based on this algorithm with the most successful foreign variants of neutralization control systems was made. Simulation for the average values of operating parameters Q = 75 m3 / h, CP = 75 g / l, and Ck = 2 g / l at minimum b = 0.02 g / l pH and maximum buffer value b = 0.47 g / l pH, and also with unidirectional extreme combination of parameters Q = 50 m3 / h, CP = 100 g / l, b = 0,02 g / l pH and Q = 100 m3 / h, CP= 50 g / l, for minimum b = 0.02 g / l pH and maximum buffering value b = 0.47 g / l pH shows that the best quality of transient processes is for the control system operating on the basis of the developed algorithm. For any combination of parameters, the transients for this control system provide better quality transients. Studies have shown that the control system based on the developed algorithm in comparison with the previously proposed systems provides better process control by reducing the time of transients and reducing the dynamic deviation of the output parameters, which improves the quality of wastewater treatment.
 Given the non-stationary process and high requirements for the cleaning parameters, manual control of this process is beyond the power of even an experienced operator. The developed mathematical model describing the dynamics of the wastewater neutralization reactor with buffer properties and the process control algorithm made it possible to proceed to the development of the control system software, which is necessary for the automated control of this process.

Continuous removal of sulfate and metals from acidic mining-impacted waters at low temperature using a sulfate-reducing bacterial consortium

The aim of this study was to develop a biological method for the simultaneous removal of sulfate and metals from acidic low-temperature mining effluents. A mixed consortium of cold-tolerant sulfate-reducing bacteria (SRB) and other microorganisms was immobilized on glass beads and exploited in an up-flow biofilm reactor for the continuous treatment of actual and synthetic mining-impacted waters (MIWs) with initial sulfate concentrations between 1580 and 5350 mg L-1. The proton acidity of the mine waters was neutralized by microbial sulfidogenesis. Metals present in the MIWs were precipitated either off-line or in-line, inside the reactor vessel. High sulfate reduction rates (SRRs), from 1000 to 4500 mg L-1 d-1 at a temperature of 11.7 ± 0.2 °C, were achieved (sulfate removal 43–87%). The bacterial consortium was found to be robust and resistant to changes in growth conditions during the bioreactor experiment. The relative abundance of SRB and the SRR increased at higher sulfate concentrations. Sulfidogenic bioreactors have the potential for treatment of acid mine drainage even at low temperature. It was demonstrated that neutral reactor conditions and high SRRs were maintained when acidic influent was fed into the reactor.

ANN model adaptation algorithm based on extended Kalman filter applied to pH control using MPC

The performance of model predictive controllers (MPCs) strongly depends on the precision of the prediction model. Nonlinear systems, such as neutralization reactors, provide special challenges to MPC design. Linear prediction models may be inadequate to describe the process at all operating points. One alternative is the use of artificial neural networks (ANNs) as prediction models. ANNs are nonlinear structures that can be trained to reproduce the process behavior. Inside MPC schemes, ANNs can rapidly predict the process response to a control action. The time-consuming step for ANN training is to obtain a representative overall data set from experiments or simulation data from the studied process. In the present work, we propose to obtain this data set from computational simulations using a first principles model. However, mismatches were found between rigorous simulation and actual pH process responses. Those deviations were naturally transferred to the internal neural model, as a consequence, actual control problems were identified. Avoiding high costs of performing actual experimental runs for ANN and MPC design, we used a real-time adaptation algorithm, based on extended Kalman filter (EKF), that acts to correct the ANN prediction while process is running. The adaptive model ANN-based MPC was able to maintain the actual controlled process, in all operating conditions tested. The sum of square error of pH was reduced in 64.3%, compared to the ANN-based MPC without model adaptation. Using a Kalman filter to adapt the internal model has significantly improved the MPC performance, reducing oscillations and maintaining the controlled variable in the setpoint, even in servo regulatory situations of industrial practice. In addition, the proposed scheme has great potential for controlling highly nonlinear processes.

Kinetic Study and Numerical Validation of the Cyanide Neutralization Process Using Alkaline Chlorination

In the recovery of precious metals from the mining industry, cyanide is the most used leaching agent due to its high efficiency. However, in some cases cyanide mining waste is dumped into bodies of water without a proper treatment, causing irreversible damage to the environment. Therefore, it is essential to have sustainable, efficient, and economic processes for the adequate treatment of cyanide toxic waste. This work presents a kinetic study of cyanide neutralization using an alkaline chlorination method on waste samples from silver leaching. Results presented demonstrate removal efficiencies of 99.7% at volumetric CN- with loading rates of about 2047 mg/l in a period of 75 h. Additionally, numerical simulation was validated considering the experimental results. For the simulation, fundamental parameters, such as mass balance, temperature and concentration, were considered to find kinetic global rate constants. The results obtained allowed to propose the design of a cyanide neutralization reactor.

Model Predictive Control Using Neural Networks [25 Years Ago

In this article, we present the application of a neural-network-based model predictive control scheme to control pH in a laboratory-scale neutralization reactor. We use a feedforward neural network as the nonlinear prediction model in an extended DMC-algorithm to control the pH-value. The training data set for the neural network was obtained from measurements of the inputs and outputs of the real plant operating with a PI-controller. Thus, no a priori information about the dynamics of the plant and no special operating conditions of the plant were needed to design the controller. The training algorithm used is a combination of an adaptive backpropagation algorithm that tunes the connection weights with a genetic algorithm to modify the slopes of the activation function of each neuron. This combination tuned out to be very robust against getting caught in local minima and it is very insensitive to the initial settings of the weights of the network.

Explicit nonlinear predictive control algorithms for Laguerre filter and sparse least square support vector machine-based Wiener model

In this paper, three computationally proficient model predictive control (MPC) algorithms for least square support vector machine (LSSVM)-based Wiener model are described. A Wiener model with Laguerre filter as dynamic linear part and LSSVM approximator as nonlinear static part is considered. Even though having excellent approximation abilities, LSSVM suffers from lack of sparseness. A pruning algorithm for LSSVM model is proposed and its comparison is made with classical pruning algorithm. The proposed pruning algorithm is able to remove 99% of support vectors with no remarkable drop in modelling accuracy. Using pruned Wiener model, three computationally efficient MPC algorithms are described. In the first algorithm, linearization of Wiener model is performed at every sampling interval and therefore control vector is determined by carrying out a quadratic optimization task. In the second algorithm, control signal is determined by an explicit control law and parameters of this control law are computed by performing lower-upper (LU) factorization of a matrix and solving linear equations without any online optimization. In the third algorithm, the parameters of explicit control law are calculated directly by another LSSVM approximator, which is trained offline. The advantages and effectiveness of proposed methods are demonstrated on the benchmark pH neutralization reactor. The control performance and computational efficiency of proposed algorithms are compared with computationally complex nonlinear MPC, which repeats a nonlinear optimization task at every sampling interval. The impact of pruning on model accuracy, computational efficiency and control accuracy is also discussed.

A Continuous Monitoring for Neutral Grounding Resistors and Reactors With Hardware Validation

Neutral grounding devices (NGDs) are used to limit ground overcurrents, control transient overvoltages, and overcome the consequent issues. Intactness of these apparatuses is vital to prevent the risk of ungrounded or solidly grounded neutral. As such, they should be continuously monitored as targeted here. This paper proposes a continuous monitoring technique that supervises the NGD impedance obtained via full-range voltage and current measurement. It benefits from a new and cost-effective voltage metering mechanism specifically designed for the studied problem. Its performance has been validated using a fabricated prototype. In contrast to existing methods, the application of the proposed technique is not limited to a specific configuration, system operation mode, or NGD. Furthermore, monitoring is performed without signal injection. This monitoring technique shows reliable operation under various conditions of the system and grounding apparatuses based on the analysis performed using PSCAD in conjunction with MATLAB.

Fast adaptive autoreclosing technique for series compensated transmission lines

Auto-reclosing renders fast restoration of the line for arcing faults which constitutes almost 80% of transmission line faults and thus improves system stability. Existing methods, using harmonics of voltage and current, have drawbacks in series-compensation environments because of associated dynamics and control mechanisms. This research work proposes a technique for adaptive autoreclosing in case of line-to-ground faults for series compensated transmission line. A technique is devised that utilises the cumulative sum of energy (CSE) calculated using the fast sparse S-transform coefficients to derive an index (PI) which discriminates between permanent and transient faults. Another extinction index (EI) is computed using CSE during the secondary arc to identify the arc extinction. This parameter is employed to alter the dead time set for the recloser and thus reclosing during fault process is avoided. The performance validation on the real-time digital simulator platform for both single- and double-circuit line with shunt compensation and neutral reactor, enhances the applicability of the proposed protection scheme for adaptive autoreclosing.

Electromagnetic Transient Analysis of 500 kV XLPE Submarine Cable Interconnection Project from Zhenhai to Zhoushan

In Zhenhai-Zhoushan transmission project, 500kV XLPE submarine cable lines are used for the first time in the world. Using electromagnetic transient simulation program(EMTP), electromagnetic transient problems in this project are investigated including electrical parameter of submarine cable lines, power frequency overvoltages, secondary-arc currents and recovery voltages during the single-phase reclosing, power frequency resonance overvoltages, induced voltages and induced currents, switching surges and so on. On the basis, the recommended configuration scheme of HV shunt reactors and neutral reactors is proposed. In addition, it is suggested to install closing resistors of 400ohm on both sides of line breakers in order to restrict switching overvoltages and ensure safe operation of submarine cable lines.

Parameter Estimation of Neutral Grounding Reactor for a Single Line to Ground Fault for Transformer

Parameter Estimation of Neutral Grounding Reactor for a Single Line to Ground Fault for Transformer

Multi-variable dependent forecast algorithm for predicting secondary arc characteristics in UHV systems with HRPC

Hybrid reactive power compensation (HRPC) combines step-controlled shunt reactors and series compensation, and will be employed in ultra-high-voltage (UHV) power systems. The single-phase auto-reclosure characteristics of secondary arcs in systems with HRPC require further investigation. In this paper, both the arc-recalling voltage and subsidiary variations in arc current are investigated with and without HRPC. The frequency components of the secondary arc current and variations in arcing time are analyzed for various influential factors, such as the neutral reactor, arc resistance, fault location, degrees of compensation of HRPC, and the length of the transmission line. The non-dominated sorting genetic algorithm II (NSGA-II) and support vector machine regression are combined to create a multi-variable dependent forecasting algorithm to predict the characteristics of the secondary arc in UHV systems with HRPC. This paper provides a theoretical reference for optimizing the parameters of HRPC, and for developing adaptive auto-reclosure schemes and protection equipment.

Practical Nonlinear Model Predictive Control Algorithm for Long Short-Term Memory Networks

Abstract This paper presents a practical nonlinear model predictive control algorithm using Long Short-Term Memory (LSTM) artificial neural networks as process model. The controller is based on the Practical Nonlinear Predictive Controller (PNMPC) algorithm, a general framework which can be used for the implementation of nonlinear model predictive controllers using nonlinear models of almost any class. LSTM has gained attention during the last years for its ability to build high fidelity models in data-driven problems, such as system identification. In the proposed control algorithm, the LSTM model is used to obtain on-line a nonlinear prediction of the system free response and a linearized version of the neural model is used to obtain a local approximation of the system step response, which is used to build the dynamic matrix of the system at each sampling instant. The proposed controller was tested in a benchmark neutralization reactor control problem. The process was identified using the LSTM approach and the identification and closed-loop results are compared with linear and Hammerstein nonlinear models. The LSTM approach outperformed the other alternatives in both identification and control problems. The results obtained in this study show that the proposed control algorithm is an option to use a state-of-the-art nonlinear MPC in a computationally affordable way.

Application of universal neutral reactor in shunt compensated transmission lines: feasibility study

Single-phase neutral reactors are used in shunt compensated transmission lines (TLs) to accelerate the extinction of the secondary arc through suppression of the arc current. Application of similar shunt reactors with the same capacities is prevalent in the transmission systems. Using the traditional method, however, for each TL, a specially designed neutral reactor is needed that could be in some cases extremely large and practically infeasible. This study investigates the feasibility of application of a properly designed neutral reactor, which is independent of the line length. This solution only depends on the ratio of positive- and zero-sequence capacitances of the line as well as the shunt reactor inductance. Accordingly, for all TLs at the same voltage level equipped with shunt reactors of the same capacities, the proposed neutral reactance would be almost identical which helps to minimise the cost and design complexity. Meanwhile, the required insulation level would be kept sufficiently small for any compensation degree. Based on the steady-state analysis and time-domain simulations conducted using the electromagnetic transient program, it is shown that for the majority of TLs, it is possible to extinguish the secondary arc within an acceptable short period of time by applying the proposed neutral reactor.

Pruning of recurrent neural models: an optimal brain damage approach

This paper considers the problem of pruning recurrent neural models of perceptron type with one hidden layer which may be used for modelling of dynamic system. In order to reduce the number of model parameters (i.e. the number of weights), the Optimal Brain Damage (OBD) pruning algorithm is adopted for the recurrent neural models. Efficiency of the OBD algorithm is demonstrated for pruning neural models of a neutralisation reactor benchmark process. For the considered neutralisation system, the OBD algorithm makes it possible to reduce as many as 60% of model parameters and reduce the validation error by some 30% when compared to the full (unpruned) models.