Adaptive Linear Model Research Articles

Incremental Learning Framework for Autonomous Robots Based on Q-Learning and the Adaptive Kernel Linear Model

The performance of autonomous robots in varying environments needs to be improved. For such incremental improvement, here we propose an incremental learning framework based on Q-learning and the adaptive kernel linear (AKL) model. The AKL model is used for storing behavioral policies that are learned by Q-learning. Both the structure and parameters of the AKL model can be trained using a novel L2-norm kernel recursive least squares (L2-KRLS) algorithm. AKL model initially without nodes and gradually accumulates content. The proposed framework allows to learn new behaviors without forgetting the previous ones. A novel local -greedy policy is proposed to speed the convergence rate of Q-learning. It calculates the exploration probability of each state for generating and selecting more important training samples. The performance of our incremental learning framework was validated in two experiments. A curve fitting example shows that the L2-KRLS based AKL model is suitable for incremental learning. The second experiment is based on robot learning tasks. The results show that our framework can incrementally learn behaviors in varying environments. Local -greedy policy-based Q-learning is faster than existing Q-learning algorithms.

Customer Satisfaction and Behaviour at Retail Outlets: an Adaptive Fuzzy Regression Model with LINGO Based Analysis

Increasing rivalry for-profit or non-profit is pushing companies to devote more and more attention to pleasing consumers with excellent quality customer services. This study aims to develop a model to analyze customer behavior in a retail store and provide accurate inference for decision making. Another critical objective for this research work is the adaptation of the faceted form of neuro-response, which is substituted by the Adaptive Fuzzy Logistic Regression Model (AFLRM). AFLRM has resulting benefits over Neuro-surface and Mean Demand Heuristic methods. A sample of 100 customers who visited or walked in the retails was used as a sample. Other than neuro-response surfaces (NRSM) and The Mean Demand Heuristic models (MDSM), the present study has accustomed a generalized form known as Adaptive Fuzzy Linear Regression Model (AFLRM) to deliver the benchmark for former models and give the highest level of accuracy for future behavior of a customer. LINGO based Markovian analysis has also been used with the above model to understand the behavior of the system under study. The significance of service and product attributes is implicitly derived via the fuzzy regression model for customer satisfaction measurement. It is observed that the critical gap between the quality of product and services and Customer Satisfaction is Product/Service Satisfaction, Motivation and Buying Experience, and Credibility and Security. The authors’ finding indicates that the effort of listening to the customer's voice should be more critical. Result analysis based on computational results concerning the questionnaire for measuring the customer behavior and the system validates the model under study. Appropriate, useful with reliable action plans for every critical product and service aspect can be developed by applying the adaptive regression methodology to control the quality of service or managing customer satisfaction, thereby providing executives with a competitive gain. Also explored the behavior of the system, i.e., whether the customer will move to the new retail outlets or they will remain in the same state by using the LINGO based software program model. Keywords: heuristic, fuzzy, Markov process, retail customer, customer behavior, LINGO, ISM.

On‐line extreme learning algorithm based identification and non‐linear model predictive controller for way‐point tracking application of an autonomous underwater vehicle

In most of the surveillance applications of autonomous underwater vehicle (AUV), very often it is intended to follow the desired horizontal way-points, where some oceanography data need to be collected. In view of this, the motion planning algorithm using way-points is investigated in this study. The proposed work involves identification of dynamics of AUV and design of adaptive model predictive controllers which includes linear adaptive model predictive controller (LAMPC) and non-linear adaptive model predictive controller (NAMPC). Owing to the fast convergence rate and robustness property, on-line sequential extreme learning machine (OS-ELM) is employed for estimating the dynamics of AUV. To improve the OS-ELM modelling performance, Jaya optimisation algorithm is applied to optimise the hidden layer parameters. The desired surveillance region is formulated in terms of way-points using heading angle obtained from desired line-of-sight path. Simulations are performed using MATLAB by applying proposed NAMPC, LAMPC and a previously reported optimal controller, namely inverse optimal self-tuning PID (IOSPID) controller. Subsequently, real-time experimentation is performed using a prototype AUV in a swimming pool. From the simulation and experimental results, it is observed that the proposed controller exhibit efficient tracking performance in face of actuator constraints as compared to LAMPC and IOSPID controller.

A Distributed Predictive Control of Energy Resources in Radiant Floor Buildings

This paper studies the impact of using different types of energy storages integrated with a heat pump for energy efficiency in radiant-floor buildings. In particular, the performance of the building energy resources management system is improved through the application of distributed model predictive control (DMPC) to better anticipate the effects of disturbances and real-time pricing together with following the modular structure of the system under control. To this end, the load side and heating system are decoupled through a three-element mixing valve, which enforces a fixed water flow rate in the building pipelines. Hence, the building temperature control is executed by a linear model predictive control, which in turn is able to exchange the building information with the heating system controller. On the contrary, there is a variable action of the mixing valve, which enforces a variable circulated water flow rate within the tank. In this case, the optimization problem is more complex than in literature due to the variable circulation water flow rate within the tank layers, which gives rise to a nonlinear model. Therefore, an adaptive linear model predictive control is designed for the heating system to deal with the system nonlinearity trough a successive linearization method around the current operating point. A battery is also installed as a further storage, in addition to the thermal energy storage, in order to have the option between the charging and discharging of both storages based on the electricity price tariff and the building and thermal energy storage inertia. A qualitative comparative analysis has been also carried out with a rule-based heuristic logic and a centralized model predictive control (CMPC) algorithm. Finally, the proposed control algorithm has been experimentally validated in a well-equipped smart grid research laboratory belonging to the ERIGrid Research Infrastructure, funded by European Union's Horizon 2020 Research and Innovation Programme.

A residual load modeling approach for household short-term load forecasting application

The household residual component of total power consumption can be considered as a portion of load demand describing the non-temperature-related factors. This component can be decomposed to irregular and predictable energy demands. The predictable component of the residual load include consumptions which are likely to have a periodic behavior. The modeling of this periodic part of the residual load can help to enhance the overall forecasting accuracy. This paper intends to model the periodic part of the residual component by capturing the behavioral patterns of overestimated and underestimated residuals as the divisions of the main residual component. Accordingly, in order to achieve its ambition, this work proposes an Adaptive Circular Conditional Expectation (ACCE) method on the basis of circular analysis to define the sub-residuals operation schedules. Consequently, an adaptive Linear Model (LM) procedure is employed to predict the residual component demand using the results of the ACCE process at each time window. Subsequently, the predicted residual is utilized to adaptively improve the performance of total electricity demand forecasting. The accuracy of the forecasting results is evaluated using Normalized Mean Absolute Error (NMAE). As a result, the proposed approach of the periodic residual demand modeling in a daily horizon leads to a promising accuracy increase of 23%. Furthermore, the proposed residual modeling method, in combination with the temperature-related component forecasting, can increase the total power consumption prediction performance by 7%. The efficacy of the proposed approach is examined via numerical analysis of real data.

Adaptive Partially Linear Regression Models by Mixing Different Estimates

This paper proposes adapting the semiparametric partial model (PLM) by mixing different estimation procedures defined under different conditions. Choosing an estimation method of PLM, from several estimation methods, is an important issue, which depends on the performance of the method and the properties of the resulting estimators. Practically, it is difficult to assign the conditions which give the best estimation procedure for the data at hand, so adaptive procedure is needed. Kernel smoothing, spline smoothing, and difference based methods are different estimation procedures used to estimate the partially linear model. Some of these methods will be used in adapting the PLM by mixing. The adapted proposed estimator is found to be a square root-consistent and has asymptotic normal distribution for the parametric component of the model. Simulation studies with different settings, and real data are used to evaluate the proposed adaptive estimator. Correlated and non-correlated regressors are used for the parametric components of the semiparametric partial model (PLM). Best results are obtained in the case of correlated regressors than in the non-correlated ones. The proposed adaptive estimator is compared to the candidate model estimators used in mixing. Best results are obtained in the form of less risk error and less convergence rate for the proposed adaptive partial linear model (PLM).

Active scanner control on paper machines

ABSTRACT The cross-directional (CD) basis weight control on paper machines is improved by optimizing the path of the scanning measurement. The optimal path results from an LQG problem and depends on how the uncertainty of the present estimate of the basis weight and the intensity of process noise vary in CD. These factors are assessed by how accurately the CD basis weight estimate predicts the measured optical transmittance with a linear adaptive model on synchronized basis weight and transmittance data. Simulations on optimized scanner path in disturbance scenarios are presented, and the practical implementation of scanner control is discussed.

An Adaptive Vehicle Sideslip Estimator for Reliable Estimation in Low and High Excitation Driving

This work proposes a novel adaptive estimator for reliable vehicle sideslip angle estimation over the full vehicle operating range, using Kalman filtering. It is shown that the vehicle state estimator with adaptive linear tire model proposed in literature only provides reliable estimation for relatively large sideslip angles. This paper proposes a new method that does not suffer this limitation. It relies on the fact that linear tire behavior is mostly a property of the tire/vehicle and to a much lesser extent of the road condition. The proposed estimator therefore contains both a fixed parameter linear tire model, and an adaptive linear tire model. The former allows for reliable and stable estimation for small sideslip angles and linear tire behavior, while the latter allows tracking of nonlinear tire behavior. A smooth transition between the tire models is obtained by adapting the corresponding model covariances in the Kalman filter according to the operating conditions. For this a measure of degree of nonlinearity in tire behavior is defined. Experimental results are provided to demonstrate the robustness and validity of the proposed approach.

Single-trial evoked potentials denoising using adaptive modelling

This study presents a method for improving the signal-to-noise ratio of single-trial event-related potentials. The method is based on adaptive linear combiner Hermite model. A variable step size least mean square algorithm is used to estimate and to adjust the parameters of the filter. The performances of the method are applied to simulated data and real event-related potential recordings. The method significantly enhances the observation of single trials and the estimation of amplitude and latency of the event-related potentials.

Single-trial evoked potentials denoising using adaptive modelling

This study presents a method for improving the signal-to-noise ratio of single-trial event-related potentials. The method is based on adaptive linear combiner Hermite model. A variable step size least mean square algorithm is used to estimate and to adjust the parameters of the filter. The performances of the method are applied to simulated data and real event-related potential recordings. The method significantly enhances the observation of single trials and the estimation of amplitude and latency of the event-related potentials.

Statistical decision methods in the presence of linear nuisance parameters and despite imaging system heteroscedastic noise: Application to wheel surface inspection

This paper proposes a novel method for fully automatic anomaly detection on objects inspected using an imaging system. In order to address the inspection of a wide range of objects and to allow the detection of any anomaly, an original adaptive linear parametric model is proposed; The great flexibility of this adaptive model offers highest accuracy for a wide range of complex surfaces while preserving detection of small defects. In addition, because the proposed original model remains linear it allows the application of the hypothesis testing theory to design a test whose statistical performances are analytically known. Another important novelty of this paper is that it takes into account the specific heteroscedastic noise of imaging systems. Indeed, in such systems, the noise level depends on the pixels’ intensity which should be carefully taken into account for providing the proposed test with statistical properties. The proposed detection method is then applied for wheels surface inspection using an imaging system. Due to the nature of the wheels, the different elements are analyzed separately. Numerical results on a large set of real images show both the accuracy of the proposed adaptive model and the sharpness of the ensuing statistical test.

Modeling Strategic Behavior in The Competitive Newsvendor Problem: An Experimental Investigation

This study experimentally investigates ordering behavior in the competitive newsvendor problem. We consider a duopoly market setting with two identical newsvendors selling the same perishable goods in a common market. Our experimental results show that average observed orders systematically deviate from the Nash equilibrium, and exhibit a similar pull‐to‐center pattern as in the classic non‐competitive newsvendor experiments: average orders fall below the Nash equilibrium in the high‐margin condition, and above the Nash equilibrium in the low‐margin condition. More importantly, the observed orders in the duopoly market are significantly higher than that in the non‐competitive newsvendor market, even in situations where standard inventory models predict no difference. We explain the ordering behavior using a strategic experience‐weighted attraction (EWA) model, which captures players' propensities for strategic thinking in game settings. Our empirical analysis suggests that the strategic EWA model generates more accurate predictions of future ordering behavior than an existing linear adaptive model without concerning strategic thinking. Further analysis shows that individuals are heterogeneous in their propensities to be a strategic player. Our research indicates the importance of modeling strategic behavior when analyzing behavioral decisions in competitive (game) environments.

Adaptive regression model for synthesizing anthropometric population data

This paper presents the development of an adaptive linear regression model for synthesizing of missing anthropometric population data based on a flexible set of known predictive data. The method is based on a conditional regression model and includes use of principal component analysis, to reduce effects of multicollinearity between selected predictive measurements, and incorporation of a stochastic component, using the partial correlation coefficients between predicted measurements. In addition, skewness of the distributions of the dependent variables is considered when incorporating the stochastic components. Results from the study show that the proposed regression models for synthesizing population data give valid results with small errors of the compared percentile values. However, higher accuracy was not achieved when the number of measurements used as independent variables was increased compared to using only stature and weight as independent variables. This indicates problems with multicollinearity that principal component regression were not able to overcome. Descriptive statistics such as mean and standard deviation values together with correlation coefficients is sufficient to perform the conditional regression procedure. However, to incorporate a stochastic component when using principal component regression requires raw data on an individual level. Relevance to industryWhen developing products, workplaces or systems, it is of great importance to consider the anthropometric diversity of the intended users. The proposed regression model offers a procedure that gives valid results, maintains the correlation between the measurements that are predicted and is adaptable regarding which, and number of, predictive measurements that are selected.

Adaptive Linear Spectral Mixture Analysis

This paper presents a theory of adaptive linear spectral mixture analysis (ALSMA), which can implement LSMA using an adaptive linear mixing model (ALMM) that adjusts and varies with spectral signatures adaptively. In doing so, a recursive LSMA (RLSMA) is developed for ALSMA to allow LSMA to update spectral signature by spectral signature without reprocessing LSMA and also to fuse LSMA results obtained by ALMM using different sets of spectral signatures. To form ALMM, the concept of RLSMA-specified virtual dimensionality is further proposed for ALSMA, which not only can find spectral signatures recursively by RLSMA to adjust ALMM but also can automatically determine the number of spectral signatures via Neyman-Pearson detection theory.

Adaptive varying-coefficient linear quantile model: a profiled estimating equations approach

We consider an estimating equations approach to parameter estimation in adaptive varying-coefficient linear quantile model. We propose estimating equations for the index vector of the model in which the unknown nonparametric functions are estimated by minimizing the check loss function, resulting in a profiled approach. The estimating equations have a bias-corrected form that makes undersmoothing of the nonparametric part unnecessary. The estimating equations approach makes it possible to obtain the estimates using a simple fixed-point algorithm. We establish asymptotic properties of the estimator using empirical process theory, with additional complication due to the nuisance nonparametric part. The finite sample performance of the new model is illustrated using simulation studies and a forest fire dataset.

Adaptive regression model for prediction of anthropometric data

This paper presents and evaluates an adaptive linear regression model for the prediction of unknown anthropometric data based on a flexible set of known predictive data. The method is based on conditional regression and includes use of principal component analysis to reduce effects of multicollinearity between the predictive variables. Results from the study show that the proposed adaptive regression model produces more accurate predictions compared to a flat regression model based on stature and weight, and also compared to a hierarchical regression model, that uses geometric and statistical relationships between body measurements to create specific linear regression equations in a hierarchical structure. An additional evaluation shows that the accuracy of the adaptive regression model increases logarithmically with the sample size. Apart from the sample size, the accuracy of the regression model is affected by the number of, and on which measurements that are, variables in the predictive dataset.

Adaptive heat pump and battery storage demand side energy management

An adaptive linear model predictive control strategy is introduced for the problem of demand side energy management, involving a photovoltaic device, a battery, and a heat pump. Moreover, the heating influence of solar radiation via the glass house effect is considered. Global sunlight radiation intensity and the outside temperature are updated by weather forecast data. The identification is carried out after adapting to a time frame witch sufficiently homogeneous weather. In this way, in spite of the linearity an increase in precision and cost reduction of up to 46% is achieved. It is validated for an open and closed loop version of the MPC problem using real data of the ambient temperature and the global radiation.

Duality in Two-Stage Adaptive Linear Optimization: Faster Computation and Stronger Bounds

In this paper we derive and exploit duality in general two-stage adaptive linear optimization models. The equivalent dualized formulation we derive is again a two-stage adaptive linear optimization model. Therefore, all existing solution approaches for two-stage adaptive models can be used to solve or approximate the dual formulation. The new dualized model differs from the primal formulation in its dimension and uses a different description of the uncertainty set. We show that the optimal primal affine policy can be directly obtained from the optimal affine policy in the dual formulation. We provide empirical evidence that the dualized model in the context of two-stage lot-sizing on a network and two-stage facility location problems solves an order of magnitude faster than the primal formulation with affine policies. We also provide an explanation and associated empirical evidence that offer insight on which characteristics of the dualized formulation make computations faster. Furthermore, the affine policy of the dual formulations can be used to provide stronger lower bounds on the optimality of affine policies.

Recurrent pattern matching based stereo image coding using linear predictors

The Multidimensional Multiscale Parser (MMP) is a pattern-matching-based generic image encoding solution which has been investigated earlier for the compression of stereo images with successful results. While first MMP-based proposals for stereo image coding employed dictionary-based techniques for disparity compensation, posterior developments have demonstrated the advantage of using predictive methods. In this paper, we focus on recent investigations on the use of predictive methods in the MMP algorithm and propose a new prediction framework for efficient stereo image coding. This framework comprises an advanced intra directional prediction model and a new linear predictive scheme for efficient disparity compensation. The linear prediction model is the main novelty of this work, combining adaptive linear models estimated by least-squares algorithm with fixed linear models provided by the block-matching algorithm. The performance of the proposed intra prediction and disparity compensation methods when applied in an MMP encoder has been evaluated experimentally. Comparisons with the current stereo image coding standards showed that the proposed MMP algorithm significantly outperforms the Stereo High Profile of H.264/AVC standard. In addition, it presents a competitive performance relative to the MV-HEVC standard. These results also suggest that current stereo image coding standards may benefit from the proposed linear prediction scheme for disparity compensation, as an extension to the omnipresent block-matching solution.

Artificial neural network based adaptive linear model predictive control

Artificial neural network based adaptive linear model predictive control