Adaptive Error Curve Learning Ensemble Model for Improving Energy Consumption Forecasting

Despite the advancement within the last decades in the field of smart grids,energy consumption forecasting utilizing the metrological features is still challenging.This paper proposes a genetic algorithm-based adaptive error curve learning ensemble(GA-ECLE)model.The proposed technique copes with the stochastic variations of improving energy consumption forecasting using a machine learning-based ensembled approach.A modified ensemble model based on a utilizing error of model as a feature is used to improve the forecast accuracy.This approach combines three models,namely CatBoost(CB),Gradient Boost(GB),and Multilayer Perceptron(MLP).The ensembled CB-GB-MLP model’s inner mechanism consists of generating a meta-data from Gradient Boosting and CatBoost models to compute the final predictions using the Multilayer Perceptron network.A genetic algorithm is used to obtain the optimal features to be used for the model.To prove the proposed model’s effectiveness,we have used a four-phase technique using Jeju island’s real energy consumption data.In the first phase,we have obtained the results by applying the CB-GB-MLP model.In the second phase,we have utilized a GA-ensembled model with optimal features.The third phase is for the comparison of the energy forecasting result with the proposed ECL-based model.The fourth stage is the final stage,where we have applied the GA-ECLE model.We obtained a mean absolute error of 3.05,and a root mean square error of 5.05.Extensive experimental results are provided,demonstrating the superiority of the proposed GA-ECLE model over traditional ensemble models.

POSITIONAL ERROR CURVES BAND OF A PLANAR LINE SEGMENT WITHIN GISs

In this paper, the positional error curve of point features was extended to an error curves band of line segment features. Firstly, the constitution and shape of the error curves band were analyzed. On this basis, the general boundary curve formula of that band was derived. Secondly, the visualizing error curves bands were realized through three exam- ples. Finally,area index has been examined by comparing numerical results from error curves band and error ellipes band.

Error Model of Curves in GIS and Digitization Experiment

A stochastic error process of curves is proposed as the error model to describe the errors of curves in GIS. In terms of the stochastic process, four characteristics concerning the local error of curves, namely, mean error function, standard error function, absolute error function, and the correlation function of errors , are put forward. The total error of a curve is expressed by a mean square integral of the stochastic error process. The probabilistic meanings and geometric meanings of the characteristics mentioned above are also discussed. A scan digitization experiment is designed to check the efficiency of the model. In the experiment, a piece of contour line is digitized for more than 100 times and lots of sample functions are derived from the experiment. Finally, all the error characteristics are estimated on the basis of sample functions. The experiment results show that the systematic error in digitized map data is not negligible, and the errors of points on curves are chiefly dependent on the curvature and the concavity of the curves.

ALGORITHM FOR SPHERICITY ERROR AND THE NUMBER OF MEASURED POINTS

The data processing technique and the method determining the optimal number of measured points are studied aiming at the sphericity error measured on a coordinate measurement machine （CMM）. The consummate criterion for the minimum zone of spherical surface is analyzed first, and then an approximation technique searching for the minimum sphericity error from the form data is studied. In order to obtain the minimum zone of spherical surface, the radial separation is reduced gradually by moving the center of the concentric spheres along certain directions with certain steps. Therefore the algorithm is precise and efficient. After the appropriate mathematical model for the approximation technique is created, a data processing program is developed accordingly. By processing the metrical data with the developed program, the spherical errors are evaluated when different numbers of measured points are taken from the same sample, and then the corresponding scatter diagram and fit curve for the sample are graphically represented. The optimal number of measured points is determined through regression analysis. Experiment shows that both the data processing technique and the method for determining the optimal number of measured points are effective. On average, the obtained sphericity error is 5.78 μm smaller than the least square solution, whose accuracy is increased by 8.63%; The obtained optimal number of measured points is half of the number usually measured.

Finite-difference time-domain modeling of curved material interfaces by using boundary condition equations method

To deal with the staircase approximation problem in the standard finite-difference time-domain（FDTD） simulation,the two-dimensional boundary condition equations（BCE） method is proposed in this paper.In the BCE method,the standard FDTD algorithm can be used as usual,and the curved surface is treated by adding the boundary condition equations.Thus,while maintaining the simplicity and computational efficiency of the standard FDTD algorithm,the BCE method can solve the staircase approximation problem.The BCE method is validated by analyzing near field and far field scattering properties of the PEC and dielectric cylinders.The results show that the BCE method can maintain a second-order accuracy by eliminating the staircase approximation errors.Moreover,the results of the BCE method show good accuracy for cylinder scattering cases with different permittivities.

Nonlinear Systems Identification via an Input-Output Model Based on a Feedforward Neural Network

This paper develops a feedforward neural network based input output model for a general unknown nonlinear dynamic system identification when only the inputs and outputs are accessible observations. In the developed model, the size of the input space is directly related to the system order. By monitoring the identification error characteristic curve, we are able to determine the system order and subsequently an appropriate network structure for systems identification. Simulation results are promising and show that generic nonlinear systems can be identified, different cases of the same system can also be discriminated by our model.