Extreme Learning Machine Optimization Research Articles

Modeling and performance evaluation of wind turbine based on ant colony optimization-extreme learning machine

In this paper, an innovative hybrid multi-variable generator’s actual-output-power predicting model is proposed based on ant colony optimization algorithm and extreme learning machine network, and a data-driven performance evaluation model is presented based on the two indices, K-means clustering algorithm and Markov chain for the performance evaluation of the wind turbines. Ant colony optimization algorithm is used to optimize the initial weights and thresholds of the extreme learning machine network, then the optimized combinations of weights and thresholds are provided into the extreme learning machine models to overcome the sensitivity problem of initialization setting and the disadvantage of easily falling into local optimum. Through the actual-output-power prediction of the WTs in a wind farm, the results show that the proposed model has more higher prediction accuracy than other methods mentioned in this paper. The optimization process also shows that the prediction accuracy is sensitive to the number of hidden-layer nodes and is relatively insensitive to other model parameters. Then, the data-driven performance evaluation models are proposed based on the error sequences obtained above. The case study is conducted and the results show that the method can evaluate the operating performance of the wind turbines correctly. The effectiveness of the evaluation results is also verified by the actual operation results.

Conditioning optimization of extreme learning machine by multitask beetle antennae swarm algorithm

Extreme learning machine (ELM) as a simple and rapid neural network has been shown its good performance in various areas. Different from the general single hidden layer feedforward neural network (SLFN), the input weights and biases in hidden layer of ELM are generated randomly, so that it only takes a little computational overhead to train the model. However, the strategy of selecting input weights and biases at random may result in ill-conditioned problems. Aiming to optimize the conditioning of ELM, we propose an effective particle swarm heuristic algorithm called Multitask Beetle Antennae Swarm Algorithm (MBAS), which is inspired by the structures of artificial bee colony (ABC) algorithm and Beetle Antennae Search (BAS) algorithm. Then, the proposed MBAS is applied for optimizing the input weights and biases of ELM to solve its ill-conditioned problems. Experiment results show that the proposed method is capable of simultaneously reducing the condition number and regression error, and achieving good generalization performance.

Classification of Tree Species at the Leaf Level based on Hyperspectral Imaging Technology

This study utilized hyperspectral imaging technology to identify eight tree species at the leaf level. The successive projections algorithm (SPA), information gain (IG), and Gini index (Gini) were used to select the feature bands. Furthermore, the binary particle swarm optimization (BPSO) algorithm was used to optimize the feature bands selected by SPA, IG, and Gini. The particle swarm optimization–extreme learning machine (PSO–ELM), linear Bayes normal classifi er (LBNC), and k-nearest neighbor (KNN) recognition models for tree species were established based on all bands, feature bands, and optimized feature bands, respectively. The experimental results show that the recognition rates of the PSO–ELM, LBNC, and KNN models based on all bands were 98.45, 99.10, and 83.67%, respectively. The SPA, IG, and Gini models can all effectively select spectral bands on tree species discrimination and greatly reduce the dimension of spectral data, in which the recognition effects of the models based on the feature bands selected by Gini were the best, and the recognition rates of the PSO–ELM, LBNC, and KNN models reached 97.55, 96.53, and 80.5%, respectively. Additionally, BPSO–SPA, BPSO–IG, and BPSO–Gini models can all further reduce the dimension of spectral data on the basis of ensuring the recognition accuracy of models, in which the models established based on the optimized feature bands selected by BPSO–Gini achieved the best recognition effect and the recognition rates of the PSO–ELM, LBNC, and KNN models reached 96.53, 96.68, and 81.05%, respectively. In general, the recognition performance of the PSO–ELM model was better than those of the LBNC and KNN models.

Power load forecasting in energy system based on improved extreme learning machine

Through the accurate prediction of power load, the start and stop of generating units in the power grid can be arranged economically and reasonably. The safety and stability of power grid operation can be maintained. First, chicken swarm optimizer based on nonlinear dynamic convergence factor (NCSO) optimizer is proposed based on chicken swarm optimizer (CSO) optimizer. In NCSO optimizer, nonlinear dynamic inertia weight and levy mutation strategy are introduced. Compared with CSO optimizer, the convergence speed and effect of NCSO optimizer are obviously improved. Second, the random parameters of extreme learning machine (ELM) model are optimized by NCSO optimizer, and NCSOELM model is established to predict the power load. Finally, the NCSO optimization extreme learning machine (NCSOELM) model is used to predict the power load, and compared with back propagation (BP), support vector machine (SVM) and CSO optimization extreme learning machine (CSOELM) model. The experimental results show that the fitting accuracy of NCSOELM model is high, and the determination coefficient r2 is above 90%. And the root mean square error value of the NCSOELM model is 0.87, 0.41, and 0.25 smaller than the root mean square error values of the support vector machine, BP, and CSOELM models, respectively. Experiments show that the model proposed in this study has high fitting effect and low prediction error, which is of positive significance for the realization of economic and safe operation of energy system.

Groundwater contamination source identification based on a hybrid particle swarm optimization-extreme learning machine

When the simulation-optimization method is applied to groundwater contamination source identification (GCSI), the numerical simulation model is usually embedded in the optimization model as a constraint condition. Hundreds and thousands of simulation model calls, while solving the optimization model, often lead to computational disaster. Establishing a surrogate model for the simulation model can effectively alleviate this drawback. When using extreme learning machine (ELM) to establish a surrogate model for a simulation model, the output weights are calculated using randomly given input weights and hidden-layer deviations, which may lead to inadequate accuracy of the surrogate model and insufficient generalization ability to samples not appearing in the training set. However, by combining the ELM with particle swarm optimization (PSO), the PSO can be used to optimize the selection of the input weights and hidden-layer deviations of the ELM; thus, calculating the output weights and establishing a surrogate model based on the PSO-ELM instead of the simulation model embedded in the optimization model to complete the GCSI. The uncertainty of the parameters is rarely considered when the simulation-optimization method is applied to GCSI. Thus, the uncertainty of the parameters is considered in GCSI based on the PSO-ELM. The results show that compared with the ELM, the PSO-ELM can establish the surrogate model with higher accuracy. The surrogate model based on PSO-ELM can be embedded in the optimization model to effectively solve GCSI problems. The predictive model based on PSO-ELM can predict the release histories of contamination sources corresponding to different random parameters.

Peak Load Forecasting for Electrical Bus based on Limited Historical Data under Complex Meteorological Condition

The historical data of peak load for electrical bus are limited and fluctuated violently. The fluctuation ability of peak load for electrical bus is nonlinear and randomly. So its prediction accuracy is low. In order to improve the accuracy of peak load forecasting for electrical bus, a peak load forecasting for electrical bus method based on limited historical data under complex weather conditions is proposed. Firstly, the influence of natural meteorology, society and other factors on peak load fluctuation for electrical bus is analysed; Secondly, based on the reduction of redundancy between features of potential feature set, the feature importance ranking is obtained by conditional mutual information (CMI). Then, according to the improved particle swarm optimization extreme learning machine suitable for small sample training, the forward feature selection is performed to determine the optimal feature subset. Finally, based on the optimal feature subset, the optimal peak load forecasting model is established.

Short-Term Traffic Flow Forecasting Based on Data-Driven Model

Short-term traffic flow forecasting is the technical basis of the intelligent transportation system (ITS). Higher precision, short-term traffic flow forecasting plays an important role in alleviating road congestion and improving traffic management efficiency. In order to improve the accuracy of short-term traffic flow forecasting, an improved bird swarm optimizer (IBSA) is used to optimize the random parameters of the extreme learning machine (ELM). In addition, the improved bird swarm optimization extreme learning machine (IBSAELM) model is established to predict short-term traffic flow. The main researches in this paper are as follows: (1) The bird swarm optimizer (BSA) is prone to fall into the local optimum, so the distribution mechanism of the BSA optimizer is improved. The first five percent of the particles with better fitness values are selected as producers. The last ten percent of the particles with worse fitness values are selected as beggars. (2) The one-day and two-day traffic flows are predicted by the support vector machine (SVM), particle swarm optimization support vector machine (PSOSVM), bird swarm optimization extreme learning machine (BSAELM) and IBSAELM models, respectively. (3) The prediction results of the models are evaluated. For the one-day traffic flow sequence, the mean absolute percentage error (MAPE) values of the IBSAELM model are smaller than the SVM, PSOSVM and BSAELM models, respectively. The experimental analysis results show that the IBSAELM model proposed in this study can meet the actual engineering requirements.

Ultra-Short-Term Wind Power Prediction by Salp Swarm Algorithm-Based Optimizing Extreme Learning Machine

Wind power generation accounts for an increasing proportion of the power grid, so efficient and accurate real-time wind power prediction is particularly important for wind power grid. In view of the strong randomness and fluctuation of wind and the difficulty of predicting wind power, a Salp Swarm Algorithms-Extremely Learning Machine (SSA-ELM) based ultra-short-term wind power prediction model is proposed. In this case, the multi-input sample set is composed of historical wind speed, temperature, wind direction, atmospheric pressure and other climatic factors that are highly correlated with wind power, and the network parameters are determined in the training process. In order to improve the adaptability and accuracy of the prediction model, the input weight matrix and hidden layer deviation of the Extreme Learning Machine (ELM) are optimized by exploring and developing the Salp Swarm Algorithm in the iterative process. Finally, the simulation experiment is conducted with the actual data of a wind farm in Henan Province, and the comparison with the traditional Extreme Learning Machine, Particle Swarm Optimization Extreme Learning Machine (PSO-ELM) and Back Propagation (BP) neural network model shows that the new method avoids falling into the local extreme value, and has faster convergence speed and higher prediction accuracy.

Dynamic Expression Recognition Based on Hybrid Features and Optimized Extreme Learning Machine Model

Facial expression recognition is an academic subject, and it can be applied not only in the field of computer vision but also in psychology and sociology. This paper proposed a robust dynamic facial expression recognition method which using hybrid features and the PSOELM (Particle Swarm Optimization-Extreme Learning Machine) model. At first, six PSOELM models were trained on the training set of six basic expressions, respectively. Then by extracting the hybrid features which fusing geometric features and texture features, the dynamic expressions were recognized by PSO-ELM models. These experiments were performed on the Cohn-Kanade+ face database. The results show that this method has a better recognition effect on dynamic facial expression recognition.

A New Probabilistic Output Constrained Optimization Extreme Learning Machine

In near decades machine learning approaches have received overwhelming attention from many researchers for solving problems that cannot be ironed out by traditional approaches. However, most of these approaches produces output that is not equivalent to the probability estimates of how credible and reliable the output can be for each prediction. One widely utilized, highly accorded for generalized performance but non-probabilistic machine learning algorithm is the Extreme Learning Machine (ELM). As with other classification systems, ELM generates outputs that cannot be treated as probabilities. Current literature shows approaches attempt to assimilate probabilistic concept in ELM however their outputs are not equivalent to probabilities. Furthermore, these methods invoke two-stage post processing procedures with iterative learning procedures which are against the salient features of ELM that highlight no iterative operations involved in learning. Hence, we want to probe in this paper the ability of ELM to produce probabilistic output from the original architecture of ELM itself while preserving the merits of ELM without the need for a post processing two-stage procedures to convert the output to probability and eliminates iterative learning to compute output weights. Two methodologies of unified probabilistic ELM framework are presented, i.e., Probabilistic Output Extreme Learning Machine (PO-ELM) and Constrained Optimization Posterior Probabilistic Outputs based Extreme Learning Machine (CPP-POELM). The proposed models are evaluated empirically on several benchmark datasets as well as real world power system applications to demonstrate its validity and efficacy in handling pattern classification problems as well as decision making process.

Global Solar Radiation Estimation and Climatic Variability Analysis Using Extreme Learning Machine Based Predictive Model

Sustainable utilization of the freely available solar radiation as renewable energy source requires accurate predictive models to quantitatively evaluate future energy potentials. In this research, an evaluation of the preciseness of extreme learning machine (ELM) model as a fast and efficient framework for estimating global incident solar radiation (G) is undertaken. Daily meteorological datasets suitable for G estimation belongs to the northern parts of the Cheliff Basin in Northwest Algeria, is used to construct the estimation model. Cross-correlation functions are applied between the inputs and the target variable (i.e., G) where several climatological information’s are used as the predictors for surface level G estimation. The most significant model inputs are determined in accordance with highest cross-correlations considering the covariance of the predictors with the G dataset. Subsequently, seven ELM models with unique neuronal architectures in terms of their input-hidden-output neurons are developed with appropriate input combinations. The prescribed ELM model’s estimation performance over the testing phase is evaluated against multiple linear regressions (MLR), autoregressive integrated moving average (ARIMA) models and several well-established literature studies. This is done in accordance with several statistical score metrics. In quantitative terms, the root mean square error (RMSE) and mean absolute error (MAE) are dramatically lower for the optimal ELM model with RMSE and MAE = 3.28 and 2.32 Wm−2 compared to 4.24 and 3.24 Wm−2 (MLR) and 8.33 and 5.37 Wm−2 (ARIMA).

Performance Sensing Data Prediction for an Aircraft Auxiliary Power Unit Using the Optimized Extreme Learning Machine.

The aircraft auxiliary power unit (APU) is responsible for environmental control in the cabin and the main engines starting the aircraft. The prediction of its performance sensing data is significant for condition-based maintenance. As a complex system, its performance sensing data have a typically nonlinear feature. In order to monitor this process, a model with strong nonlinear fitting ability needs to be formulated. A neural network has advantages of solving a nonlinear problem. Compared with the traditional back propagation neural network algorithm, an extreme learning machine (ELM) has features of a faster learning speed and better generalization performance. To enhance the training of the neural network with a back propagation algorithm, an ELM is employed to predict the performance sensing data of the APU in this study. However, the randomly generated weights and thresholds of the ELM often may result in unstable prediction results. To address this problem, a restricted Boltzmann machine (RBM) is utilized to optimize the ELM. In this way, a stable performance parameter prediction model of the APU can be obtained and better performance parameter prediction results can be achieved. The proposed method is evaluated by the real APU sensing data of China Southern Airlines Company Limited Shenyang Maintenance Base. Experimental results show that the optimized ELM with an RBM is more stable and can obtain more accurate prediction results.

Study on Icing Prediction of Power Transmission Lines Based on Ensemble Empirical Mode Decomposition and Feature Selection Optimized Extreme Learning Machine

The ice coating on the transmission line is extremely destructive to the safe operation of the power grid. Under natural conditions, the thickness of ice coating on the transmission line shows a nonlinear growth trend and many influencing factors increase the difficulty of forecasting. Therefore, a hybrid model was proposed in this paper, which mixed Ensemble Empirical Mode Decomposition (EEMD), Random Forest (RF) and Chaotic Grey Wolf Optimization-Extreme Learning Machine (CGWO-ELM) algorithms to predict short-term ice thickness. Firstly, the Ensemble Profit Mode Decomposition model was introduced to decompose the original ice thickness data into components representing different wave characteristics and to eliminate irregular components. In order to verify the accuracy of the model, two transmission lines in ‘hunan’ province were selected for case study. Then the reserved components were modeled one by one, building the random forest feature selection algorithm and Partial Autocorrelation Function (PACF) to extract the feature input of the model. At last, a component prediction model of ice thickness based on feature selection and CGWO-ELM was established for prediction. Simulation results show that the model proposed in this paper not only has good prediction performance, but also can greatly improve the accuracy of ice thickness prediction by selecting input terminal according to RF characteristics.

A Real-Time BOD Estimation Method in Wastewater Treatment Process Based on an Optimized Extreme Learning Machine

It is difficult to capture the real-time online measurement data for biochemical oxygen demand (BOD) in wastewater treatment processes. An optimized extreme learning machine (ELM) based on an improved cuckoo search algorithm (ICS) is proposed in this paper for the design of soft BOD measurement model. In ICS-ELM, the input weights matrices of the extreme learning machine and the threshold of the hidden layer are encoded as the cuckoo’s nest locations. The best input weights matrices and threshold are obtained by using the strong global search ability of improved cuckoo search algorithm. The optimal results can be used to improve the precision of forecasting based on less number of neurons of the hidden layer in ELM. Simulation results show that the soft sensor model has good real-time performance, high prediction accuracy, and stronger generalization performance for BOD measurement of the effluent quality compared to other modeling methods such as back propagation (BP) network in most cases.

Development of online classification system for construction waste based on industrial camera and hyperspectral camera.

Construction waste is a serious problem that should be addressed to protect environment and save resources, some of which have a high recovery value. To efficiently recover construction waste, an online classification system is developed using an industrial near-infrared hyperspectral camera. This system uses the industrial camera to capture a region of interest and a hyperspectral camera to obtain the spectral information about objects corresponding to the region of interest. The spectral information is then used to build classification models based on extreme learning machine and resemblance discriminant analysis. To further improve this system, an online particle swarm optimization extreme learning machine is developed. The results indicate that if a near-infrared hyperspectral camera is used in conjunction with an industrial camera, construction waste can be efficiently classified. Therefore, extreme learning machine and resemblance discriminant analysis can be used to classify construction waste. Particle swarm optimization can be used to further enhance the proposed system.

Effect of high-humidity hot air impingement blanching (HHAIB) and drying parameters on drying characteristics and quality of broccoli florets

In current work, the influences of high-humidity hot air impingement blanching (HHAIB) time (60, 90, 120, and 150 s), drying temperature (60, 65, 70, and 75 °C), and air velocity (6, 9, and 12 m/s) on drying characteristics and quality attributes of broccoli florets were explored. Extreme learning machine (ELM) was employed to describe the drying behavior of broccoli florets. Results showed that proper HHAIB pretreatment can extensively increase drying rate compared to the control group (unblanched samples). The entire drying process of broccoli florets occurred in the falling period. Besides, proper HHAIB pretreatment can enhance vitamin C preservation, the color quality, and the rehydration capacity of dried broccoli florets. Based on error analysis results, the prediction accuracy of the optimal ELM model with 4-50-1 topology is found to be satisfied for the moisture ratio prediction of broccoli florets during air impingement drying process, with the R2, , and MSE reached to 0.9993, 8.04e−5, and 2.01e−4, respectively.

Purity detection of goat milk based on electronic tongue and improved artificial fish swarm optimized extreme learning machine

The nutritional value of goat milk is higher than that of cow milk, which is scarce and expensive. Therefore, many bad credit enterprises mix cow’s milk into goat milk and sell it as pure goat milk, which infringes on the rights and interests of consumers. The adulteration of dairy products has a long history. Many techniques have been applied for the detection of dairy products’ adulteration, but no obvious results have been achieved. The electronic tongue system, based on a virtual instrument, can provide fast detection of goat milk’s quality. The qualitative identification of different purities of goat milk by kernel principal component analysis (KPCA) has achieved 100% accuracy. By comparing the four algorithms of the optimal extreme learning machine (ELM), it was determined that the quantitative model for the prediction of adulterated goat milk, established by the improved artificial fish swarm optimized ELM method, had high prediction accuracy. The prediction set coefficient R2 was 0.998, the average absolute error (MAE) was 0.083, and the root mean square error (MSE) was 0.011. According to its unique advantages, the electronic tongue provides a new idea and method for the detection of food adulteration. As a modern intelligent sensory instrument, the electronic tongue has great potential for brand identification as well as the identification of goat milk and goat milk powder adulteration.

Forecasting China’s Natural Gas Consumption Based on AdaBoost-Particle Swarm Optimization-Extreme Learning Machine Integrated Learning Method

With the orderly advancement of ‘China’s energy development strategic action plan’, the natural gas industry has achieved unprecedented development. Currently, it is planned that by 2020, China’s natural gas consumption will account for at least 10% of the total primary energy consumption, have an orderly and improved energy structure, and achieved energy-saving and emission-reduction targets. Therefore, the accurate prediction of natural gas consumption becomes significantly important. Firstly, based on the research status of forecasting methods and the factors which affect natural gas consumption, this paper used the particle swarm optimization (PSO) algorithm to obtain the input layer weight, and used the optimized extreme learning machine (ELM) algorithm to obtain the hidden layer threshold; by using PSO-ELM as the base predictor and the AdaBoost algorithm, we have constructed the natural gas consumption integrated learning prediction model. Secondly, from the perspective of different provinces and industries, we deeply analyze the current status of natural gas consumption, and the random forest algorithm is used to extract the core influencing factors of natural gas consumption as the independent variables of the prediction model. Finally, data on China’s natural gas consumption from 1995 to 2017 are selected, then the feasibility analysis and comparative analysis with other methods are performed. The results show: 1) Using the random forest algorithm to extract the core influencing factors, economic growth, population, household consumption and import dependence degree are significantly representative. 2) Based on the AdaBoost integrated learning algorithm, transforming the weak predictor with poor prediction effect into a strong predictor with strong prediction effect, compared with PSO-ELM, AdaBoost-ELM and ELM algorithm, with R-Square as 0.9999, Mean Square Error (MSE) as 0.8435, Mean Absolute Error (MAE) as 0.2379, Mean Absolute Percentage Error (MAPE) as 0.0008, effectively validated the significant effect of the AdaBoost-PSO-ELM prediction model. 3) Based on the AdaBoost-PSO-ELM prediction model, predict the natural gas core influencing factors and natural gas consumption in the year of 2018–2030. There is an apparent growth trend in the next 13 years, and the average growth rate of natural gas consumption has reached 7.68%.

Research and Application of a Hybrid Wind Energy Forecasting System Based on Data Processing and an Optimized Extreme Learning Machine

Accurate wind speed forecasting plays a significant role for grid operators and the use of wind energy, which helps meet increasing energy needs and improve the energy structure. However, choosing an accurate forecasting system is a challenging task. Many studies have been carried out in recent years, but unfortunately, these studies ignore the importance of data preprocessing and the influence of numerous missing values, leading to poor forecasting performance. In this paper, a hybrid forecasting system based on data preprocessing and an Extreme Learning Machine optimized by the cuckoo algorithm is proposed, which can overcome the limitations of the single ELM model. In the system, the standard genetic algorithm is added to reduce the dimensions of the input and utilize the time series model for error correction by focusing on the optimized extreme learning machine model. And according to screened results, the 5% fractile and 95% fractile are applied to compose the upper and lower bounds of the confidence interval, respectively. The assessment results indicate that the hybrid system successfully overcomes some limitations of the single Extreme Learning Machine model and traditional BP and Mycielski models and can be an effective tool compared to traditional forecasting models.

Adaptive reactionless control strategy via the PSO-ELM algorithm for free-floating space robots during manipulation of unknown objects

The operation of free-floating space robots causes significant challenges to the attitude control of a spacecraft body, resulting from the strong dynamic coupling existing in the system as well as the appending unknown property from a captured object. Conventional reactionless motion planning and control methods require the precise knowledge of system dynamics with ideal path tracking performance, which are impossible to achieve in reality. To solve this problem, an adaptive reactionless path planning and control integrated strategy is proposed for free-floating space robots while manipulating unknown objects. For details, a SW-RLS (Slide-windowed Recursive Least Square) algorithm is employed to construct the adaptive reactionless path planning by identifying and compensating the unknown property in the form of a time-variable matrix for the momentum equation of the system. Simultaneously, a robust adaptive control strategy via the PSO-ELM (Particle Swarm Optimization-Extreme Learning Machine) algorithm is proposed to track the dynamic change in the planned path, based on the foundation of a designed PD controller for reducing the error. With computational efficiency, an adaptive controller via PSO-ELM aims at learning and compensating for the unknown property intelligently; then, a robust controller is used to reject the model uncertainties. The notable feature of the proposed strategy is that it requires neither an accurate system model nor any information about the unknown property. Most importantly, the design can dynamically realize reactionless path tracking performance. Finally, the simulations of a free-floating space robot capturing an unknown object and comparisons with existing methods are presented to demonstrate the effectiveness of the proposed scheme.