Stochastic Configuration Networks Research Articles

Modular stochastic configuration network based prediction model for NOx emissions in municipal solid waste incineration process

The accurate prediction of the nitrogen oxides (NOx) emissions is extremely important for pollutant control in municipal solid waste incineration (MSWI) process. Modular neural network (MNN) provides a support for predicting NOx emissions to overcome the limitations of a single model in nonlinear processes and different climatic conditions. However, the design of sub model for MNN is a challenge. We propose a method based on MNN and adaptive ensemble stochastic configuration network (m-AESCN) in this work. First, fuzzy C-means algorithm decomposes the task into several sub datasets with similar characteristics, and an evaluation function is used to guarantee the optimal decomposition result. Second, for each sub dataset, an adaptive ensemble stochastic configuration network as the sub model of modular neural network. The optimal output for sub model can be obtained by an adaptive weighting method based on variance. Third, in testing phase, a sub model activation method based on cluster centroid is proposed to select a suitable sub model. Besides, the m-AESCN method is validated by the real data of an MSWI plant and shows considerable performance in two different datasets. And the proposed method is compared with several modeling methods, such as stochastic configuration network, support vector regression, random forest and so on. The experimental results under two datasets prove that the maximum improvements in accuracy of m-AESCN are 39.74% and 31.67%, respectively.

Modeling Method for Aerobic Zone of A2O Based on KPCA-PSO-SCN

Sewage treatment plants face significant problems as a result of the annual growth in urban sewage discharge. Substandard sewage discharge can also be caused by rising sewage treatment expenses and unpredictable procedures. The most widely used sewage treatment process in urban areas is the Anaerobic–Anoxic–Oxic (A2O) sewage treatment process. Therefore, modeling the sewage treatment process and predicting the effluent quality are of great significance. A process modeling method based on Kernel Principal Component Analysis–Particle Swarm Optimization–Stochastic Configuration Network (KPCA-PSO-SCN) is proposed for the A2O aerobic wastewater treatment process. Firstly, eight auxiliary variables were determined through mechanism analysis, including Chemical Oxygen Demand (COD) and ammonia nitrogen (NH4+) and nitrate nitrogen (NO3−) of influent water, pH, temperature (T), Mixed Liquor Suspended Solid (MLSS), Dissolved Oxygen (DO) and hydraulic residence time (HRT) in the aerobic zone. Dimensionality reduction was carried out using the kernel principal component analysis method based on the Gaussian function, and the eight-dimensional data were changed to five-dimensional data, which improved the running speed and efficiency of subsequent models. Then, according to the advantages of the particle swarm optimization algorithm, such as low calculation cost and fast convergence, combined with the advantages of stochastic configuration network general approximation performance, the PSO-SCN model was established to predict the three water quality indexes of effluent COD, NH4+, and NO3− for the aerobic zone. The experimental results proved the effectiveness of the model. Compared with classic water quality prediction algorithm models such as SCN, PSO-BP, RBF, PSO-RBF, etc., the superiority of the PSO-SCN algorithm model was demonstrated.

Adaptive learning-based recoil control for deepwater drilling riser systems

This paper deals with the adaptive learning based recoil suppression control problem of deepwater drilling riser systems subject to parametric perturbations, platform heave motion, and friction resistance force of drilling fluid discharge. First, a stochastic configuration network based approximator is presented to predict the friction resistance force of drilling fluid discharge. Then, model reference adaptive learning recoil controllers are developed for the drilling riser system, and the existence condition and supporting algorithm are provided. Simulation results show that: (i) the designed stochastic configuration network approximator is efficient to predict the friction resistance of drilling fluid discharge on the riser; (ii) the presented model reference adaptive learning recoil controllers are effective to restrain the recoil response of the riser; (iii) compared with the existing recoil controllers, which are all based on the exact dynamic recoil control models, the model reference adaptive learning recoil controllers proposed in this paper are no longer constrained by the accuracy of recoil model; (iv) the model reference adaptive learning recoil controllers are more robust than the existing ones to the unmodeled dynamics, systematic uncertainty, and parametric perturbations, and have better transient performance of system thereby guaranteeing the safety of the drilling riser systems effectively.

KPCA-WPHM-SCNs-based remaining useful life prediction method for motor rolling bearings

Motor rolling bearings are the important supporting components of motors. It can ensure the stable operation of motor equipment in the power grid, and bearing life prediction of it is a key issue. To solve the problem of low accuracy of remaining useful life (RUL) prediction for motor rolling bearings, a neural network model based on Weibull proportional hazards model (WPHM) and stochastic configuration networks (SCNs) is proposed. To better extract and analyze features of the bearing vibration signal in both time and frequency domains, kernel principal component analysis (KPCA) is used to reduce the dimensionality of the data. Then, a WPHM model using the top three contributing feature parameters is built, which sets the start time based on the failure rate curve and reliability function. Finally, the validity of the model is verified with the rolling bearing full life cycle dataset from the IEEE PHM 2012 Data Challenge, and a comparison with other machine learning models shows that the accuracy of the proposed model in RUL prediction is higher.

Adaptive stochastic configuration network ensemble for structural reliability analysis

This paper develops a new adaptive surrogate model method for structural reliability analysis, which is constructed by approximating the limit state function (LSF) using stochastic configuration network (SCN) ensemble model. The proposed ensemble model integrates three SCN base models and sets weights for each prediction result, reducing outliers and greatly improving the output accuracy and stability of the algorithm. In constructing a surrogate model for reliability analysis, a new adaptive sampling strategy is designed, including a novel learning function. The new strategy enables the most informative sample points are selected and significantly reduce the number of required sample point. Without further calls to the LSF, the structural failure probability is quickly evaluated using Monte Carlo simulation in terms of the constructed SCN ensemble surrogate model. Three benchmark problems, including a highly nonlinear problem with small failure probability and two multi-variable engineering structural problems, are used to verify the effectiveness of the proposed method.

Energy Consumption Prediction of Steelmaking Process Based on Improved Whale Optimization Algorithm and Stochastic Configuration Network

Energy Consumption Prediction of Steelmaking Process Based on Improved Whale Optimization Algorithm and Stochastic Configuration Network

Activity recognition in rehabilitation training based on ensemble stochastic configuration networks

Activity recognition in rehabilitation training based on ensemble stochastic configuration networks

Ultra-short-term power load forecasting method based on stochastic configuration networks and empirical mode decomposition

Ultra-short-term power load forecasting (USTPLF) can provide strong support and guarantee the decisions on unit start-up, shutdown, and power adjustment. The ultra-short-term power load (USTPL) has strong non-smoothness and nonlinearity, and the time-series characteristics of the load data themselves are difficult to explore. Therefore, to fully exploit the intrinsic features of the USTPL, a stochastic configuration networks (SCNs) USTPLF method based on K-means clustering (K-means) and empirical mode decomposition (EMD) is proposed. First, the load data are decomposed into several intrinsic mode functions (i.e., IMFs) and residuals (i.e., Res) by EMD. Second, the IMFs are classified by K-means, and the IMF components of the same class are summed. Third, the SCNs is used to forecast the electric load on the basis of the classified data. Lastly, on the basis of the real load of Shenzhen City, the proposed method is applied for emulation authentication. The result verifies the efficiency of the proposed measure.

Soft sensing of calcination zone temperature of lime rotary kiln based on principal component analysis and stochastic configuration networks

The combustion state of the lime rotary kiln is mainly reflected in the temperature of the calcination zone. However, due to the high temperature in the calcination zone during the production process, it is not possible to directly measure using instruments. This paper proposes a soft sensing strategy to address the issue of difficulty in measuring the temperature of the calcination zone of a lime rotary kiln. Firstly, due to significant disturbances and coupling of production process variables, principal component analysis is used to reduce the input dimension and computational complexity of the data, thereby improving the efficiency of the soft sensing model. Then, stochastic configuration networks with good predictive performance is used to model and predict the principal components. Meanwhile, an improved Hammersley-based Salp swarm algorithm is proposed to optimize the key parameter regularization factor of the stochastic configuration networks. Finally, the validity of the soft sensing model is verified by the actual collected production data of the lime rotary kiln. Multiple evaluation indicators and comparison results indicate that the proposed soft sensing model for the calcination zone temperature of lime rotary kilns has higher prediction accuracy and modeling quality compared to other models.

Distributed sparse learning for stochastic configuration networks via alternating direction method of multipliers

Distributed sparse learning for stochastic configuration networks via alternating direction method of multipliers

Real-time reliability analysis of micro-milling processes considering the effects of tool wear

The development of micro-milling technology has promoted the high-precision machining of complex and small-size parts in modern manufacturing. Accurate cutting force control and reasonable machining parameters selection are conducive to the sustainable machining of micro-milling. In this work, a micro-milling mechanical model considering the change of tool wear, tool runout, and chip separation status with cutting time is proposed with the improved Gated Recurrent Unit with Multi-objective Dandelion Optimizer and mathematical modeling. The modeling accuracy is verified by experiments on Al6061. Combined with the established framework and the practical physical information, the posterior distribution of process parameters is inverted using the Bayesian updating method. Finally, the reliability calculation of machining system parameters is efficiently realized utilizing the posterior information and the High-dimensional Model Representation with the Stochastic configuration network method. The influence degree of the change of cutting parameters on the extreme value of micro-milling force is evaluated. The research results can provide analytical methods and theoretical guidance to ensure safe processing.

A stochastic configuration networks based on Harris hawks optimizer

Stochastic configuration networks (SCNs), as a novel incremental generation model with supervisory mechanism, has an excellent superiority in solving large-scale data regression and classification problems. However, the accuracy of the SCNs is affected by the assignation and selection of some network parameters significantly. Harris hawk optimizer (HHO) algorithm is a new meta-heuristic algorithm that simulates the foraging and anti-predation behavior of Harris hawk. In this paper, a SCNs based on HHO algorithm is first introduced, termed as HHO-SCNs. As the performance of SCNs is related to regularization parameter r and scale factor lambda of weights and biases, then HHO is employed to give better parameters for SCNs automatically. A numerical function and six benchmark datasets are used to verify the regression performance of the proposed model. Three benchmark datasets are introduced to illustrate the effectiveness of the proposed model for classification performance. Experimental results demonstrate the feasibility and validity of HHO-SCNs compared with incremental random vector functional link, SCNs, fast SCNs, and SCNs based on whale optimization algorithm. The proposed HHO-SCNs improves the generalization performance of standard SCNs, and provides a new idea for expanding the development and application of SCNs.

Incremental Multilayer Broad Learning System With Stochastic Configuration Algorithm for Regression

Broad learning system (BLS) is a novel randomized learning framework which has a faster modeling efficiency. Although BLS with incremental learning has a better extendibility for updating model rapidly, the incremental mode of BLS lacks self-supervision mechanism which cannot adjust the structure adaptively. Learning from the idea of stochastic configuration network (SCN), a novel incremental multi-layer broad learning system based on stochastic configuration (SC) algorithm is proposed for regression, termed as IMLBLS-SC. Firstly, to improve the feature learning ability, SC algorithm is adopted to configure the parameters of enhancement nodes instead of random weights. Secondly, the multi-layer model with enhancement nodes can be added gradually according to supervision mechanism without human intervention. Thirdly, all the enhancement nodes and feature nodes are fully connected with output nodes. Finally, 2 function approximation problems and 8 classical datasets are selected to verify the regression performance of IMLBLS-SC, experimental results demonstrate that IMLBLS-SC outperforms the RVFLN, SCN, BLS and BSCN.

A regularized stochastic configuration network based on weighted mean of vectors for regression.

The stochastic configuration network (SCN) randomly configures the input weights and biases of hidden layers under a set of inequality constraints to guarantee its universal approximation property. The SCN has demonstrated great potential for fast and efficient data modeling. However, the prediction accuracy and convergence rate of SCN are frequently impacted by the parameter settings of the model. The weighted mean of vectors (INFO) is an innovative swarm intelligence optimization algorithm, with an optimization procedure consisting of three phases: updating rule, vector combining, and a local search. This article aimed at establishing a new regularized SCN based on the weighted mean of vectors (RSCN-INFO) to optimize its parameter selection and network structure. The regularization term that combines the ridge method with the residual error feedback was introduced into the objective function in order to dynamically adjust the training parameters. Meanwhile, INFO was employed to automatically explore an appropriate four-dimensional parameter vector for RSCN. The selected parameters may lead to a compact network architecture with a faster reduction of the network residual error. Simulation results over some benchmark datasets demonstrated that the proposed RSCN-INFO showed superior performance with respect to parameter setting, fast convergence, and network compactness compared with other contrast algorithms.

Low-voltage series arc fault detection based on ECMC and VB-SCN

The currents of low-voltage series arc faults are non-stationary and nonlinear, so the selection of fault features and the adaptability of the detection algorithm will affect the detection results. In response to these issues, firstly, a feature selection method combining Euclidean distance, classifier criterion, max-relevance min-redundancy and clustering index is proposed, ECMC. This method can filter out weak correlation features and redundant information in the time, frequency and time–frequency domains. The optimal feature data set constructed by ECMC can fully reflect the characteristics of arc faults. Then, a stochastic configuration network (SCN) based on variational Bayesian (VB) optimization, VB-SCN, is proposed. This method uses VB to optimize the output weight and scale function of SCN iteratively. VB-SCN can improve the quality of hidden-layer nodes and the generalization ability of the network while ensuring the adaptive learning of current features, making arc fault detection fast and stable. Finally, the experiments based on the self-built sample set prove that the optimal feature dataset constructed by ECMC has perfect separability. VB-SCN can adaptively realize the accurate and fast detection of arc faults. The detection accuracy is stable between 98.2506% and 100%, which has better robustness and stability.

Digital twin-based thermal analysis of a frequency synthesizer

A frequency synthesizer is a key component of radars and plays an important role in radar radio frequency transceiver systems. In this study, a digital twin (DT) model of the virtual entity of a frequency synthesizer is proposed to perform thermal analysis. First, the virtual entity of the DT model is established according to the physical entity of the frequency synthesizer. Then, based on the designed DT model of the virtual entity of the frequency synthesizer, the temperature distribution of the model and cooling channel is analyzed, and the temperature distribution of the key components is obtained. Finally, using the results from the DT virtual entity model, a frequency synthesizer temperature prediction method based on a stochastic configuration network (SCN) algorithm is proposed. In the test phase, the characteristic and virtual temperature vectors are used as the input and output of the SCN training set, respectively. Then, 10 different sets of inputs are selected to obtain 10 sets of simulated temperatures and SCN predicted temperatures. The results indicate that the SCN algorithm based on the DT frequency synthesizer can accurately predict the temperature of the physical frequency synthesizer. The thermal analysis of the frequency synthesizer based on the DT provides accurate solutions and reduces the time required by physical components to adjust to high temperatures.

A nonlinear industrial soft sensor modeling method based on locality preserving stochastic configuration network with utilizing unlabeled samples

Stochastic configuration network (SCN) is an emerging incremental randomized regression modeling technology with the advantages of adaptively determining the hidden layer parameters, and has been successfully applied to industrial soft sensor modeling field. However, the traditional SCN model is intrinsically a supervised learner, which has the underlying assumption that all the training samples are labeled. In fact, most of process samples are unlabeled and the labeled samples are relatively rare in real industrial scenarios. To handle this issue, this paper presents one modified SCN model, called locality preserving SCN (LPSCN), for semi-supervised industrial soft sensor modeling. In this method, all the training samples, including the labeled and the unlabeled, are fed into the soft sensor model, where the labeled samples are used to capture the modeling error, while the unlabeled samples help construct the local adjacency graph. Based on these two kinds of samples, the supervised optimization objective in the traditional SCN is improved to be a semi-supervised version by minimizing the modeling error and preserving the local data relationship simultaneously. Furthermore, the random parameter configuration mechanism is deduced under the modified semi-supervised optimization framework. A new inequality constraint condition with considering the unlabeled samples is obtained to generate the hidden layer nodes incrementally so that the LPSCN model structure is determined automatically. Experiments on two real industrial systems demonstrate that the proposed LPSCN method outperforms the SCN method in terms of the soft sensor prediction performance.

Fault diagnosis of oil-immersed transformer based on MGTO-BSCN

In order to improve the accuracy and reliability of fault diagnosis of oil-immersed power transformers, a fault diagnosis method based on the Modified Artificial Gorilla Troops Optimizer (MGTO) and the Stochastic Configuration Networks with Block Increments (BSCN) is proposed. First, the original artificial gorilla troop optimization algorithm is improved, which effectively improves the convergence speed and optimization accuracy of the algorithm. Secondly, the conventional Stochastic Configuration Networks (SCN) learning methodology is modified when the fault diagnosis model is constructed. The original SCN adopts point incremental approach to gradually add hidden nodes, while BSCN adopts block increment approach to learn features. It significantly accelerates training. MGTO algorithm is used to jointly optimize regularization parameter and scale factor in BSCN model, and the fault diagnosis model with the highest accuracy is constructed. The experimental results show that the accuracy of MGTO-BSCN for transformer fault diagnosis reaches 95.9%, which is 3.5%, 9.9% and 11.7% higher than BSCN fault diagnosis models optimized by GTO, Grey Wolf Optimizer (GWO) and Particle Swarm Optimization (PSO) respectively, reflecting the superiority of MGTO algorithm. Meanwhile, the comparison with the traditional model shows that the proposed method has obvious advantages in diagnostic effect.

Multitarget Stochastic Configuration Network and Applications

Existing stochastic configuration network (SCN)-based modeling methods are underperformed in handling multitarget regression problems. An important reason is that they ignore the intertarget correlations, which have an important effect on improving the modeling accuracy. To enhance the performance of these SCN-based models, in this article, a novel multitarget SCN (MTSCN) modeling approach is presented. The <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">L</i> <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2,1</sub> norm of a structure matrix can be utilized to explicitly reveal the correlations between multiple targets and an <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">L</i> <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sub> term are attached to the cost function of SCN. Considering the nonsmoothness of the constructed cost function, an alternating optimization algorithm is adopted to compute the structure matrix and the output weights of MTSCN. Then, a new supervisory mechanism is proposed to ensure the convergence of MTSCN. Finally, experimental results using the synthetic data and several real-world datasets show that the developed MTSCN is more superior to other modeling methods in resolving multitarget modeling problems.

Modeling and parameter identification of verification systems using stochastic configuration networks

The gas flow verification system with positive pressure sonic nozzle is calling for high requirements on the stability of gas source pressure. It is difficult to guarantee the stagnation cabin pressure be controlled within its targeted range, because of varying boundary conditions. To solve the above problems, this paper presents a modeling method based on mechanism and data-driven for gas flow verification system with the positive pressure sonic nozzle. The stochastic configuration network is used to dynamically identify parameters in the mechanism model. After substituting the identification results into the model, the RMSE of the pressure and temperature in the stagnation cabin obtained from the simulation and the actual data are 1.26 × 104 and 2.64 × 104, respectively. Considering the dead zone of the regulating valve and the replacement of regulating valve in the actual operation, the model is analyzed by disturbance experiment. Experimental results show that the dynamic mathematical model could reflect the changing trend of the pressure in the pressure regulating process and prove its effectiveness.