The prediction of peak power load is a critical factor directly impacting the stability of power supply, characterized significantly by its time series nature and intricate ties to the seasonal patterns in electricity usage. Despite its crucial importance, the current landscape of power peak load forecasting remains a multifaceted challenge in the field. This study aims to contribute to this domain by proposing a method that leverages a combination of three primary models - the GRU model, self-attention mechanism, and Transformer mechanism - to forecast peak power load. To contextualize this research within the ongoing discourse, it’s essential to consider the evolving methodologies and advancements in power peak load forecasting. By delving into additional references addressing the complexities and current state of the power peak load forecasting problem, this study aims to build upon the existing knowledge base and offer insights into contemporary challenges and strategies adopted within the field. Data preprocessing in this study involves comprehensive cleaning, standardization, and the design of relevant functions to ensure robustness in the predictive modeling process. Additionally, recognizing the necessity to capture temporal changes effectively, this research incorporates features such as “Weekly Moving Average” and “Monthly Moving Average” into the dataset. To evaluate the proposed methodologies comprehensively, this study conducts comparative analyses with established models such as LSTM, Self-attention network, Transformer, ARIMA, and SVR. The outcomes reveal that the models proposed in this study exhibit superior predictive performance compared to these established models, showcasing their effectiveness in accurately forecasting electricity consumption. The significance of this research lies in two primary contributions. Firstly, it introduces an innovative prediction method combining the GRU model, self-attention mechanism, and Transformer mechanism, aligning with the contemporary evolution of predictive modeling techniques in the field. Secondly, it introduces and emphasizes the utility of “Weekly Moving Average” and “Monthly Moving Average” methodologies, crucial in effectively capturing and interpreting seasonal variations within the dataset. By incorporating these features, this study enhances the model’s ability to account for seasonal influencing factors, thereby significantly improving the accuracy of peak power load forecasting. This contribution aligns with the ongoing efforts to refine forecasting methodologies and addresses the pertinent challenges within power peak load forecasting.

Recently several researchers have proposed various methods to build intelligent stock trading and portfolio management systems using rapid advancements in artificial intelligence including machine learning techniques. However, existing technical analysis-based stock price prediction studies primarily depend on price change or price-related moving average patterns, and information related to trading volume is only used as an auxiliary indicator. This study focuses on the effect of changes in trading volume on stock prices and proposes a novel method for short-term stock price predictions based on trading volume patterns. Two rapid volume decrease patterns are defined based on the combinations of multiple volume moving averages. The dataset filtered using these patterns is learned through the supervised learning of neural networks. Experimental results based on the data from Korea Composite Stock Price Index and Korean Securities Dealers Automated Quotation, show that the proposed prediction system can achieve a trading performance that significantly exceeds the market average.

This paper describes a fast and effective algorithm for refining the parameter estimates of multicomponent third-order polynomial phase signals (PPSs). The efficiency of the proposed algorithm is accompanied by lower signal-to-noise ratio (SNR) threshold, and computational complexity. A two-step procedure is used to estimate the parameters of multicomponent third-order PPSs. In the first step, an initial estimate for the phase parameters can be obtained by using fast Fourier transformation (FFT), k-means algorithm and three time positions. In the second step, these initial estimates are refined by a simple moving average filter and singular value decomposition (SVD). The SNR threshold of the proposed algorithm is lower than those of the non-linear least square (NLS) method and the estimation refinement method even though it uses a simple moving average filter. In addition, the proposed method is characterized by significantly lower complexity than computationally intensive NLS methods. Simulations confirm the effectiveness of the proposed method.

There are many web sites where net users can post and distribute their illustration images. A typical way to draw a digital illustration is first to draw rough lines on a paper and then to trace the lines on a graphics-tablet by hand. The input lines usually contain fluctuation due to hand-drawing, which limits the quality of illustration. Therefore, it is important to remove the fluctuation and to smooth the lines while maintaining sharp features such as corners. Although naive applications of moving average filters can smooth input lines, they may cause over-smoothing artifacts in which sharp features are lost by the filtering. This paper describes an improved line smoothing method using adaptive moving averages, which smoothes input lines while keeping high curvature points. The proposed method evaluates curvatures of input lines and adaptively controls the filter-size to reduce the over-smoothing artifacts. Experiments demonstrated advantages of the proposed method over the previous method in terms of achieving smoothing effect while still preserving sharp feature preservation.

In this paper, we propose a new method to estimate and effectively reduce the effect of thermal asperity (TA) in the perpendicular magnetic recording (PMR) channels with the state trellis. The TA is estimated from the state trellis, then its average is used to modify the equalized signal entering the Viterbi detector. For the partial response (PR) targets with DC component, the proposed method with a maximum-likelihood detector can improve the bit error rate performance by more than an order of magnitude when TA occurs and degrades when the giant magneto-resistive (GMR) nonlinearity and base line wander (BLW) effects are present. Unlike the previous studies, this method allows the use of PR targets with DC component under the presence of TA.

OFDM (Orthogonal Frequency Division Multiplexing) is widely used in wideband wireless communication systems due to its excellent performance. One of the most important operations in OFDM receivers is preamble detection. This paper addresses a general form of extended differential detection methods, which is a combination of differential detection and a moving average filter. This paper also presents a filter size determination method that achieves satisfactory performance in various channel environments.

In this paper the transfer function of a system with windowed current integration is discussed. This kind of integration is usually used in a sampling mixer and the current is generated by a transconductance amplifier (TA). The parasitic capacitance (Cp) and the output resistance of the TA (Ro,TA) before the sampling mixer heavily affect the performance. Calculations based on a model including the parasitic capacitance and the output resistance of the TA is carried out. Calculation results show that due to the parasitic capacitance, a notch at the sampling frequency appears, which is very harmful because it causes the gain near the sampling frequency to decrease greatly. The output resistance of the TA makes the depth of the notches shallow and decreases the gain near the sampling frequency. To suppress the effect of Cp and Ro,TA, an operational amplifier is introduced in parallel with the sampling capacitance (Cs). Simulation results show that there is a 17 dB gain increase while Cs is 1,pF, gm is 9,mS, N is 8 with a clock rate of 800,MHz.

Ultrasonic distance measurement using the pulse-echo method is based on the determination of the time of flight of ultrasonic waves. The pulse-compression technique, in which the cross-correlation function of a detected ultrasonic wave and a transmitted ultrasonic wave is obtained, is the conventional method used for improving the resolution of distance measurement. However, the calculation of a cross-correlation operation requires high-cost digital signal processing. This paper presents a new method of sensor signal processing within the pulse-compression technique using a delta-sigma modulated single-bit digital signal. The proposed sensor signal processing method consists of a cross-correlation operation employing single-bit signal processing and a smoothing operation involving a moving average filter. The proposed method reduces the calculation cost of the digital signal processing of the pulse-compression technique.

This paper proposes an adaptive channel estimation scheme, which uses different moving average length and pilot gain for different mobile environments. It is based on MSE method and extensive simulations under various environments for WCDMA physical layer. The scheme applies a computationally efficient and easily implemented pilot filter on WCDMA forward channel. For different mobile channel environments, the optimal combination of moving average length and pilot gain for low SNR is achieved. The simulation results illustrate that the adaptive scheme can achieve much lower BER compared with two other adaptive schemes, especially when the speed of mobile user is high. And the BER performance of the proposed scheme is insensible to the mobile speed.

In this paper we propose a subsequence matching algorithm that supports moving average transform of arbitrary order in time-series databases. Moving average transform reduces the effect of noise and has been used in many areas such as econometrics since it is useful in finding the overall trends. The proposed algorithm extends the existing subsequence matching algorithm proposed by Faloutsos et al. (SUB94 in short). If we applied the algorithm without any extension, we would have to generate an index for each moving average order and would have serious storage and CPU time overhead. In this paper we tackle the problem using the notion of index interpolation. Index interpolation is defined as a searching method that uses one or more indexes generated for a few selected cases and performs searching for all the cases satisfying some criteria. The proposed algorithm, which is based on index interpolation, can use only one index for a pre-selected moving average order k and performs subsequence matching for arbitrary order m ( k). We prove that the proposed algorithm causes no false dismissal. The proposed algorithm can also use more than one index to improve search performance. The algorithm works better with smaller selectivities. For selectivities less than 10-2, the degradation of search performance compared with the fully-indexed case--which is equivalent to SUB94--is no more than 33.0% when one index is used, and 17.2% when two indexes are used. Since the queries with smaller selectivities are much more frequent in general database applications, the proposed algorithm is suitable for practical situations.

This paper investigates the modeling of non-linearity on the generation of the single trial evoked potential signal (s-EP) by means of using a mixed radial basis function neural network (M-RBFN). The more emphasis is put on the contribution of spontaneous EEG term to s-EP signal. The method is based on a nonlinear M-RBFN neural network model that is trained simultaneously with the different segments of EEG/EP data. Then, the output of the trained model (estimator) is a both fitted and reduced (optimized) nonlinear model and then provide a global representation of the passage dynamics between spontaneous brain activity and poststimulus periods. The performance of the proposed neural network method is evaluated using a realistic simulation and applied to a real EEG/EP measurement.