Reduction Of Computational Load Research Articles

Overcoming challenges in crash prediction modeling using discretized duration approach: An investigation of sampling approaches

Until recently, statistical approaches used for real-time crash prediction modeling were limited to case-control design and “sampling of alternatives” approaches. A recent study has developed a duration-based real-time crash prediction model capable of incorporating dynamic (time-varying) covariates within its framework. The modeling approach discretizes the duration between crashes into equal time intervals which can be modeled as alternatives in a multinomial logit framework. The approach, however, requires a reformulation of the original crash dataset to fit its modeling framework which results in considerably large data making model estimation computationally demanding. Additionally, validation of the model in the original study is based on crash data from just one interstate, I-405, assuming homogenous highway segments each 5 miles in length. This study improves upon the original study by investigating sampling techniques that can be applied to the reformulated data to reduce computational load using 2019 crash data from two interstates, I-40 and I-55, in Memphis, Tennessee. Furthermore, discretization of inter-crash duration is undertaken following non-homogenous segmentation of the interstates that is based on highway geometry, terrain, and posted speed limit. To accomplish the study objectives, a relatively small future window of 1 h with 15-minute time intervals is used to discretize the inter-crash duration and obtain the reformulated data. Sampling of crashes for model estimation is then done at the crash, epoch, and segment levels to answer the question of which sampling technique (by crash, epoch, or segment) would result in reasonable accuracy when compared with the complete (100%) data. Results show that 25% of samples drawn at the epoch level can result in a considerable reduction of computational load while providing reasonably consistent estimates.

Hybrid domain decomposition filters for advection–diffusion PDEs with mobile sensors

This paper examines one aspect of computational control, namely the reduction of computational load for the real-time estimation of distributed parameter systems over large spatial domains. One way to improve the performance of the state estimator is to employ a small number of mobile devices that are free to move within the spatial domain while collecting process information. Another way is to use numerical methods to reduce the computational load as dictated by the discretization index of the finite dimensional representation of the filter. This is done via the use of domain decomposition methods and the domain decomposition filter is presented here for the first time along with its well-posedness. In the subdomain surrounding the sensor a refined grid is used as it improves the spatial resolution, and in the subdomain excluding the vicinity of the sensing device, a naïve observer with coarse grid is used which results in a significant reduction of the computational cost. However, due to the transmission conditions that couple the information between the inner subdomain surrounding the sensor and the outer subdomain excluding the sensor, the filters in the two subdomains implement an information exchange protocol rendering them consensus filters. The two approaches are combined thus resulting in a state estimation scheme where the mobile sensor residing in the inner subdomain carries its own refined grid. Not only is the sensor mobile, but also the refined grid surrounding the sensor is moving along with the sensor. A further reduction in computational costs is presented in the form of a hybrid DD filter wherein the filter kernel is rendered sparse thus ensuring that no residual measurement signals enter as disturbance signals in the naïve observer defined over the outer subdomain. Extensive simulation studies of a 2D PDE with realistic parameters are included to demonstrate the proposed hybrid DD filter.

Model predictive consensus in networked autonomous systems using discrete Laguerre functions and event triggering approach

This paper presents the consensus problem of general linear discrete time networked autonomous agents, using distributed model predictive control (DMPC) and event triggering algorithm. The computational complexity of conventional DMPC is high due to the large values of prediction horizon and control horizon required for real time systems. The increased computational load becomes an obstruction in the practical implementation of DMPC. To solve the issue, a novel design methodology of parameterizing the control input using discrete Laguerre functions is adopted in this paper. The predictive control protocol is applied to a group of networked autonomous agents operating on event triggered consensus algorithm. The proposed event triggered algorithm offers the advantages of lesser communication between the agents along with reduction in controller updates. The closed loop stability of the networked system is established using Lyapunov criterion, based on the terminal state’s equality constraint. Comparison of the proposed algorithm with existing works through numerical examples shows significant reduction in computational load, design parameters and data transmission, while achieving comparable performance. To validate the theoretical results, simulation is carried out on a group of autonomous vehicles and the usefulness of the proposed algorithm on system performance is illustrated.

Vibration control of semi-active suspension by the neural network that learned the optimal preview control of MLD model

Advances in image processing technology have made it possible to measure the surface shape of the road ahead while driving. We propose a new semi-active suspension control method that considers the forward road surface shape. A vehicle model equipped with a semi-active suspension can be expressed as an MLD (Mixed Logical Dynamical) model. When the shape of the road ahead can be measured accurately, that the information on the future disturbances is available before the vehicle undergoes. In this paper, we formulate the finite time optimization problem of the MLD model in consideration of the future disturbances as a MIQP (Mixed Integer Quadratic Programming) problem in the same way as the conventional optimal control problem without future disturbance. The solution to the MIQP problem can be obtained by a commonly available solver software. However, because the MIQP problem is classified as NP-hard, it is hard to obtain the control action within the control cycle period required by the vibration control problem with general computers for vehicles. In this paper, we achieve this reduction in computational load by constructing an approximation function for the designed controller. A multilayer neural network is adopted for the approximation. The performance evaluation of the proposed control method was evaluated by a simulation. In the simulation study, the proposed method was able to achieve better ride comfort with the equivalent suspension stroke compared to the traditional MLD predictive control and the Skyhook approximation methods. Moreover, the proposed method enables to generate the control signal within the control cycle period.

A Selective Ensemble Approach for Accuracy Improvement and Computational Load Reduction in ANN-Based PV Power Forecasting

Day-ahead power forecasting is an effective way to deal with the challenges of increased penetration of photovoltaic power into the electric grid, due to its non-programmable nature. This is significantly beneficial for smart grid and micro-grids application. Machine learning and hybrid approaches are well assessed techniques, able to provide effective forecasting with a data-driven approach based on previous measurements from existing power plants. Ensemble methods can be employed to increase solar power forecasting accuracy, by running several independent forecasting models in parallel. In this paper, a novel selective approach is proposed and assessed, where independently trained neural networks are evaluated in terms of accuracy, in order to properly select a suitable forecasting. Moreover, in order to reduce the associated computational burden, suitably developed new normalization approaches are proposed and evaluated. The considered experimental case study shows that the combination of the proposed procedures is able to increase accuracy and to mitigate the overall computational load, resulting in a simple and lightweight algorithm. Additionally, a comparison with other commonly used techniques has shown that the proposed approach is robust with respect to dataset limited size and discontinuities.

Edge-Computing-Assisted Virtual Reality Computation Offloading: An Empirical Study

Offloading heavy virtual reality (VR) computational operations to a network edge computation entity is receiving increasing attention as a tool to wirelessly and energy efficiently provide low-end client devices with high-quality and immersive interactive VR services anytime and anywhere across the globe. In this work, we aim to provide an understanding of various characteristics of VR computation offloading through comprehensive experiments conducted using a prototype testbed for edge-assisted VR processing and streaming. First, we investigate the benefits of VR offloading in terms of computational load and power consumption reduction for a client device compared to standalone operation. Next, we measure VR traffic patterns, including frame size and data and packet rates with various settings, such as different resolution and encoding options. We also measure several performance metrics associated with the quality of experience, namely, frame rate, packet loss rate, and image quality, with various configuration settings. Then, we present latency measurement studies and investigate per-component latency with various settings. Furthermore, we report the rigorous experiments performed to study the impacts of latency and motion patterns on the black borders formed due to image reprojection and the overfilling technique used to eliminate these black borders.

Inference Skipping for More Efficient Real-Time Speech Enhancement With Parallel RNNs

Deep neural network (DNN) based speech enhancement models have attracted extensive attention due to their promising performance. However, it is difficult to deploy a powerful DNN in real-time applications because of its high computational cost. Typical compression methods such as pruning and quantization do not make good use of the data characteristics. In this paper, we introduce the Skip-RNN strategy into speech enhancement models with parallel RNNs. The states of the RNNs update intermittently without interrupting the update of the output mask, which leads to significant reduction of computational load without evident audio artifacts. To better leverage the difference between the voice and the noise, we further regularize the skipping strategy with voice activity detection (VAD) guidance, saving more computational load. Experiments on a high-performance speech enhancement model, dual-path convolutional recurrent network (DPCRN), show the superiority of our strategy over strategies like network pruning or directly training a smaller model. We also validate the generalization of the proposed strategy on two other competitive speech enhancement models.

Three-Dimensional Human Head Reconstruction Using Smartphone-Based Close-Range Video Photogrammetry

Creation of head 3D models from videos or pictures of the head by using close-range photogrammetry techniques has many applications in clinical, commercial, industrial, artistic, and entertainment areas. This work aims to create a methodology for improving 3D head reconstruction, with a focus on using selfie videos as the data source. Then, using this methodology, we seek to propose changes for the general-purpose 3D reconstruction algorithm to improve the head reconstruction process. We define the improvement of the 3D head reconstruction as an increase of reconstruction quality (which is lowering reconstruction errors of the head and amount of semantic noise) and reduction of computational load. We proposed algorithm improvements that increase reconstruction quality by removing image backgrounds and by selecting diverse and high-quality frames. Algorithm modifications were evaluated on videos of the mannequin head. Evaluation results show that baseline reconstruction is improved 12 times due to the reduction of semantic noise and reconstruction errors of the head. The reduction of computational demand was achieved by reducing the frame number needed to process, reducing the number of image matches required to perform, reducing an average number of feature points in images, and still being able to provide the highest precision of the head reconstruction.

Training and inference for integer-based semantic segmentation network

Semantic segmentation has been a major topic in research and industry in recent years. However, due to the computation complexity of pixel-wise prediction and backpropagation algorithm, semantic segmentation has been demanding in computation resources, resulting in slow training and inference speed and large storage space to store models. Existing schemes that speed up segmentation network change the network structure and come with noticeable accuracy degradation. However, neural network quantization can be used to reduce computation load while maintaining comparable accuracy and original network structure. Semantic segmentation networks are different from traditional deep convolutional neural networks (DCNNs) in many ways, and this topic has not been thoroughly explored in existing works. In this paper, we propose a new quantization framework for training and inference of segmentation networks, where parameters and operations are constrained to 8-bit integer-based values for the first time. Full quantization of the data flow and the removal of square and root operations in batch normalization give our framework the ability to perform inference on fixed-point devices. Our proposed framework is evaluated on mainstream semantic segmentation networks like FCN-VGG16 and DeepLabv3-ResNet50, achieving comparable accuracy against floating-point framework on ADE20K dataset and PASCAL VOC 2012 dataset.

The reduction of computation load in SBAS system during the polynomial approximation of observations

Processing of the observations from GNSS augmentation system on an example of SBAS system is considered in the study. To compute the corrections the SBAS control center collects and processes enormous amount of data, some of the applied matrix algorithms of estimation of the polynomial coefficients are computationally expensive. Instead of the estimation of the polynomial coefficients with the matrix operations the representation of polynomial value by means of a linear combination of previous observations is suggested in the study. Significant advantage in terms of computation time (up to 250 times) for cycle-slip detection procedure is demonstrated for presented method which is important for aviation because of strict integrity requirements in avia-space sector.

Jointly Optimizing Offloading Decision and Bandwidth Allocation with Energy Constraint in Mobile Edge Computing Environment

Edge computing is regarded as the new paradigm to provide cloud computing capacity for edge users. Meanwhile, more and more edge devices are connected to edge servers. The end devices choose to offload their computation task to the edge server for reducing computation load and improving task process efficiency. However, due to the limited communication capacity of the edge base station, the edge server needs to reasonably assign bandwidth resources to improve the quality of service (QoS). In this paper, we focus on the offloading problem of the partial computation task. The objective is to reduce the task computation time by reasonably allocating the bandwidth resource and making a moderate task offloading proportion. Firstly, the optimization problems for users and edge servers are successively discussed. Then, the game theory is adopted and the Stackelberg game model is built. Furthermore, the existence of the Nash equilibrium is proven. Meanwhile, the computation algorithm is designed. Finally, the numerical simulation is conducted to evaluate the performance of the proposed algorithm. The results imply that the performance of the proposed algorithm is better than that of the benchmark algorithms in terms of the task computation time and energy consumption.

An efficient block Gauss–Seidel iteration method for the space fractional coupled nonlinear Schrödinger equations

In this paper, we propose an efficient block Gauss–Seidel iteration method for solving the complex linear equations arising from the space fractional coupled nonlinear Schrödinger (CNLS) equations. The proposed method avoids the inverse of coefficient matrix when solving linear equations, which can greatly reduce computation load and storage space. Furthermore, the convergence of the iteration method is proved theoretically and the number of iteration steps is estimated. Numerical results are presented to show the block Gauss–Seidel iteration method can be quite competitive when compared with others.

Reliable Time Propagation Algorithms for PMF and RBPMF.

This paper addresses the reliable time propagation algorithms for Point Mass Filter (PMF) and Rao–Blackwellized PMF (RBPMF) for the nonlinear estimaton problem. The conventional PMF and RBPMF process the probability diffusion for the time propagation with the direct sampled-values of the process noise. However, if the grid interval is not dense enough, it fails to represent the statistical characteristics of the noise accurately so the performance might deteriorate. To overcome that problem, we propose time propagation convolution algorithms adopting Moment Matched Gaussian Kernel (MMGK) on regular grids through mass linear interpolation. To extend the dimension of the MMGK that can accurately describe the noise moments up to the kernel length, we propose the extended MMGK based on the outer tensor product. The proposed time propagation algorithms using one common kernel through the mass linear interpolation not only improve the performance of the filter but also significantly reduce the computational load. The performance improvement and the computational load reduction of the proposed algorithms are verified through numerical simulations for various nonlinear models.

Analysis method based on coupled heat transfer and CFD simulations for buildings with thermally complex building envelopes

A thermal environment simulation method was developed for a structure with a thermally complex building envelope. The method used a heat transfer simulation (HTS), computational fluid dynamics (CFD), and a two-dimensional heat-flow calculation tool called Hygrabe2D. The general HTS does not support the modeling of buildings with a radial shape. Therefore, the building envelope was evaluated using a two-dimensional heat-flow computation tool coupled to the HTS. In the HTS, the amount of advection between zones for buildings with volumes that should be divided into multiple zones was unknown. The indoor surface temperature calculated by Hygrabe2D was coupled to the HTS, and the advection and convective heat transfer coefficients between zones were calculated via CFD, which were passed to the HTS. The accuracy of the proposed method was verified and validated. The influence of calculation accuracy was investigated by comparing the accuracies in the presence and absence of the coupling of the convective heat transfer coefficient. The initial conditional dependence of coupled advection between zones was confirmed, although it had no significant effect on the calculation accuracy. Through verification of the computational load reduction, we achieved computational accuracy with a small number of couplings. The proposed method was used to evaluate the thermal environment of a building by calculating the annual room temperature and temperature inside the building skin.

Factor Graph-Based Smoothing Without Matrix Inversion for Highly Precise Localization

We consider the problem of localizing a manned, semi-autonomous, or autonomous vehicle in the environment using information coming from the vehicle's sensors, a problem known as navigation or simultaneous localization and mapping (SLAM) depending on the context. To infer knowledge from sensors' measurements, while drawing on a priori knowledge about the vehicle's dynamics, modern approaches solve an optimization problem to compute the most likely trajectory given all past observations, an approach known as smoothing. Improving smoothing solvers is an active field of research in the SLAM community. Most work is focused on reducing computation load by inverting the involved linear system while preserving its sparsity. The present paper raises an issue which, to the knowledge of the authors, has not been addressed yet: standard smoothing solvers require explicitly using the inverse of sensor noise covariance matrices. This means the parameters that reflect the noise magnitude must be sufficiently large for the smoother to properly function. When matrices are close to singular, which is the case when using high precision modern inertial measurement units (IMU), numerical issues necessarily arise, especially with 32-bits implementation demanded by most industrial aerospace applications. We discuss these issues and propose a solution that builds upon the Kalman filter to improve smoothing algorithms. We then leverage the results to devise a localization algorithm based on fusion of IMU and vision sensors. Successful real experiments using an actual car equipped with a tactical grade high performance IMU and a LiDAR illustrate the relevance of the approach to the field of autonomous vehicles.

A modified hybrid finite‐element method for a semi‐implicit mass‐based non‐hydrostatic kernel

AbstractA hybrid finite‐element vertical discretization method for a semi‐implicit mass‐based non‐hydrostatic kernel is a feasible high‐order discretization approach which combines the advantages of both finite‐differential and finite‐element methods. In this article, we put forward a modified version of the existing hybrid finite‐element method to reduce computation load and improve precision. A key feature of our modified method is the presence, in the enlarging step, of new levels which are not equally spaced, but rather based on Gaussian quadrature. A higher‐order accuracy may be thus achieved with less enlarged levels by virtue of the properties of Gaussian quadrature. The modified method is also designed to fulfil the constraints required by the dynamic kernel, which themselves are crucial to ensure stability. A set of 2D and 3D test cases are conducted so as to confirm the accuracy and the stability of the new method.

Analysis of Unified Power Flow Controller Steady-State Power Flow Regulation Capability and Its Key Factors

As the latest generation of flexible AC transmission system (FACTS) devices, the unified power flow controller (UPFC) has comprehensive control capability and has played an important role in modern power systems. Research on UPFC steady-state power flow regulation capability is valuable for the design and operation of UPFC projects. To analyze the UPFC regulation capability, in this paper, the interaction between the UPFC and the external system is studied and the key factors of UPFC regulation capability are analyzed. It is demonstrated that the changes in the voltage magnitude and phase angle difference of UPFC-embedded line terminals will hinder the UPFC from regulating the power flow and decrease the UPFC regulation capability. Therefore, the UPFC power flow regulation capability is not only related to the parameters of the UPFC and its installation line, but also related to grid parameters. To analyze the UPFC active power flow regulation capability in practical power grids, the relationship between the UPFC regulation capability and grid parameters is deduced, and an estimation method is proposed to calculate the regulation range. Then, the estimation method is applied in the Chuxiong Power Grid and compared with power flow calculations. The results verify the UPFC power flow regulation characteristics given by the analysis and the effectiveness of the estimation method. Besides, it is shown that the estimation method can lower the requirements for the power flow calculation program, and reduce computation load.

A novel gateway-based solution for remote elderly monitoring.

Internet of Things (IoT) technologies have been applied to various fields such as manufacturing, automobile industry and healthcare. IoT-based healthcare has a significant impact on real-time remote monitoring of patients' health and consequently improving treatments and reducing healthcare costs. In fact, IoT has made healthcare more reliable, efficient, and accessible. Two major drawbacks which IoT suffers from can be expressed as: first, thelimited battery capacityof thesensorsis quickly depleted due to the continuous stream of data; second, the dependence of the system on the cloud for computations and processing causes latency in data transmission which is not accepted in real-time monitoring applications. This research is conducted to develop a real-time, secure, and energy-efficient platform which provides a solution for reducing computation load on the cloud and diminishing data transmission delay. In the proposed platform, the sensors utilize a state-of-the-art power saving technique known as Compressive Sensing (CS). CS allows sensors to retrieve the sensed data using fewer measurements by sending a compressed signal. In this framework, the signal reconstruction and processing are computed locally on a Heterogeneous Multicore Platform (HMP) device to decrease the dependency on the cloud. In addition, a framework has been implemented to control the system, set different parameters, display the data as well as send live notifications to medical experts through the cloud in order to alert them of any eventual hazardous event or abnormality and allow quick interventions. Finally, a case study of the system is presented demonstrating the acquisition and monitoring of the data for a given subject in real-time. The obtained results reveal that the proposed solution reduces 15.4% of energy consumption in sensors, that makes this prototype a good candidate for IoT employment in healthcare.

An Optimal Filter for Singular Systems With Stochastic Multiplicative Disturbance

This brief proposes an optimal filtering algorithm for linear singular systems with stochastic multiplicative noise by using the second restricted equivalent transformation (SRET) to reduce computation load, avoid curse of dimensionality and improve precision. The presented filter can be guarantee optimum in the sense of linear minimum mean-square error (LMMSE) by employing the projection theorem. A performance analysis and simulation results verify the proposed algorithm.

Computational load reduction of the agent guidance problem using Mixed Integer Programming.

This paper employs a solution to the agent-guidance problem in an environment with obstacles, whose avoidance techniques have been extensively used in the last years. There is still a gap between the solution times required to obtain a trajectory and those demanded by real world applications. These usually face a tradeoff between the limited on-board processing performance and the high volume of computing operations demanded by those real-time applications. In this paper we propose a deferred decision-based technique that produces clusters used for obstacle avoidance as the agent moves in the environment, like a driver that, at night, enlightens the road ahead as her/his car moves along a highway. By considering the spatial and temporal relevance of each obstacle throughout the planning process and pruning areas that belong to the constrained domain, one may relieve the inherent computational burden of avoidance. This strategy reduces the number of operations required and increases it on demand, since a computationally heavier problem is tackled only if the simpler ones are not feasible. It consists in an improvement based solely on problem modeling, which, by example, may offer processing times in the same order of magnitude than the lower-bound given by the relaxed form of the problem.