Short-term electrical load forecasting plays an important role in the safety, stability, and sustainability of power production scheduling process. An accurate prediction can provide a reliable decision for system management. To solve limitation existing methods dealing with time-series data, causing poor stability non-ideal accuracy, this paper proposed attention-based encoder-decoder network Bayesian optimization to do short-term forecasting. Proposed model is based on architecture gated...

Prediction based on time series has a wide range of applications. Due to the complex nonlinear and random distribution data, performance learning prediction models can be reduced by modeling bias or overfitting. This paper proposes novel planar flow-based variational auto-encoder model (PFVAE), which uses long- short-term memory network (LSTM) as designs (VAE) data predictor overcome noise effects. In addition, internal structure VAE is transformed using flow, enables it learn fit...

Due to the nonlinear modeling capabilities, deep learning prediction networks have become widely used for smart agriculture. Because sensing data has noise and complex nonlinearity, it is still an open topic improve its performance. This paper proposes a Reversible Automatic Selection Normalization (RASN) network, integrating normalization renormalization layer evaluate select module of model. The accuracy been improved effectively by scaling translating input with learnable parameters....

Compared with mechanism-based modeling methods, data-driven based on big data has become a popular research field in recent years because of its applicability. However, it is not always better to have more when building forecasting model practical areas. Due the noise and conflict, redundancy, inconsistency time-series data, accuracy may reduce contrary. This paper proposes deep network by selecting understanding improve performance. Firstly, self-screening layer (DSSL) maximal information...

The environment and development are major issues of general concern. After much suffering from the harm environmental pollution, human beings began to pay attention protection started carry out pollutant prediction research. A large number air predictions have tried predict pollutants by revealing their evolution patterns, emphasizing fitting analysis time series but ignoring spatial transmission effect adjacent areas, leading low accuracy. To solve this problem, we propose a network with...

Air quality plays a vital role in people’s health, and air forecasting can assist decision making for government planning sustainable development. In contrast, it is challenging to multi-step forecast accurately due its complex nonlinear caused by both temporal spatial dimensions. Deep models, with their ability model strong nonlinearities, have become the primary methods forecasting. However, because of lack mechanism-based analysis, uninterpretability makes decisions risky, especially when...

Based on the collected weather data from agricultural Internet of Things (IoT) system, changes in can be obtained advance, which is an effective way to plan and control sustainable production. However, it not easy accurately predict future trend because always contain complex nonlinear relationship with multiple components. To increase prediction performance precision agriculture IoT this study used a deep learning predictor sequential two-level decomposition structure, were decomposed into...

Smart agricultural sensing has enabled great advantages in practical applications recently, making it one of the most important and valuable systems. For outdoor plantation farms, prediction climate data, such as temperature, wind speed, humidity, enables planning control production to improve yield quality crops. However, is not easy accurately predict trends because data are complex, nonlinear, contain multiple components. This study proposes a hybrid deep learning predictor, which an...

Air pollution (mainly PM2.5) is one of the main environmental problems about air quality. prediction and early warning a prerequisite for prevention control. However, it not easy to accurately predict long-term trend because collected PM2.5 data have complex nonlinearity with multiple components different frequency characteristics. This study proposes hybrid deep learning predictor, in which are decomposed into by empirical mode decomposition (EMD) firstly, convolutional neural network (CNN)...

The control effect of various intelligent terminals is affected by the data sensing precision. filtering method has been typical soft computing used to promote level. Due difficult recognition practical system and empirical parameter estimation in traditional Kalman filter, a neuron-based filter was proposed paper. Firstly, framework improved designed, which neuro units were introduced. Secondly, functions excavated with nonlinear autoregressive model. optimized process reduce unpractical...

Weather is an essential component of natural resources that affects agricultural production and plays a decisive role in deciding the type production, planting structure, crop quality, etc. In field agriculture, medium- long-term predictions temperature humidity are vital for guiding activities improving yield quality. However, existing intelligent models still have difficulties dealing with big weather data predicting applications, such as striking balance between prediction accuracy...

It is crucial to predict PM2.5 concentration for early warning regarding and the control of air pollution. However, accurate prediction has been challenging, especially in long-term prediction. monitoring data comprise a complex time series that contains multiple components with different characteristics; therefore, it difficult obtain an by single model. In this study, integrated predictor proposed, which original are decomposed into three components, is, trend, period, residual then...

Mobility is a significant robotic task. It the most important function when robotics applied to domains such as autonomous cars, home service robots, and underwater vehicles. Despite extensive research on this topic, robots still suffer from difficulties moving in complex environments, especially practical applications. Therefore, ability have enough intelligence while key issue for success of robots. Researchers proposed variety methods algorithms, including navigation tracking. To help...

Trend prediction based on sensor data in a multi-sensor system is an important topic. As the number of sensors increases, we can measure and store more data. However, increase has not effectively improved performance. This paper focuses this problem presents distributed predictor that overcome unrelated noise: First, define causality entropy to calculate measurement’s causality. Then, series coefficient (SCC) proposed select high causal measurement as input To traditional deep learning...

To solve the problem of traversal multi-target path planning for an unmanned cruise ship in unknown obstacle environment lakes, this study proposed a hybrid algorithm. The algorithm can be divided into two parts. First, was transformed traveling salesman problem, and improved Grey Wolf Optimization (GWO) used to calculate sequence. GWO optimized convergence factor by introducing Beta function, which improve speed traditional Second, based on planned target sequence, D* Lite implement between...

The prediction information has effects on the emergency prevention and advanced control in various complex systems. There are obvious nonlinear, nonstationary, complicated characteristics time series. Moreover, multiple variables time‐series impact each other to make more difficult. Then, a solution of for multivariate was explored this paper. Firstly, compound neural network framework designed with primary auxiliary networks. attempted extract change features series as well interactive...

ABSTRACT Traditional motion models often cannot describe real‐world systems accurately when using the Kalman filter (KF) for target tracking. This paper aims to achieve an adaptive estimation of states and proposes a KF coupled with neural networks (NNs). First, framework is proposed state recognition tracking, which couples different NN classical KF. Second, filtering algorithm introduced. utilizes NNs learn patterns total Gaussian probability density sequence performs iterative updates...