Nonlinear inverse modeling of sensor based on back-propagation fuzzy logical system

Objective To correct the nonlinear error of sensor output,a new approach to sensor inverse modeling based on Back-Propagation Fuzzy Logical System(BP FS) is presented.Methods The BP FS is a computationally efficient nonlinear universal approximator,which is capable of implementing complex nonlinear mapping from its input pattern space to the output with fast convergence speed.Results The neuro-fuzzy hybrid system,i.e.BP FS,is then applied to construct nonlinear inverse model of pressure sensor.The experimental results show that the proposed inverse modeling method automatically compensates the associated nonlinear error in pressure estimation,and thus the performance of pressure sensor is significantly improved.Conclusion The proposed method can be widely used in nonlinearity correction of various kinds of sensors to compensate the effects of nonlinearity and temperature on sensor output.

Dynamic ocean inverse modeling based on differentiable rendering

Learning and inferring underlying motion patterns of captured 2D scenes and then re-creating dynamic evolution consistent with the real-world natural phenomena have high appeal for graphics and animation.To bridge the technical gap between virtual and real environments,we focus on the inverse modeling and reconstruction of visually consistent and property-verifiable oceans,taking advantage of deep learning and differentiable physics to learn geometry and constitute waves in a self-supervised manner.First,we infer hierarchical geometry using two networks,which are optimized via the differentiable renderer.We extract wave components from the sequence of inferred geometry through a network equipped with a differentiable ocean model.Then,ocean dynamics can be evolved using the reconstructed wave components.Through extensive experiments,we verify that our new method yields satisfactory results for both geometry reconstruction and wave estimation.Moreover,the new framework has the inverse modeling potential to facilitate a host of graphics applications,such as the rapid production of physically accurate scene animation and editing guided by real ocean scenes.

Determination of urban runoff coefficient using time series inverse modeling

Runoff coefficient is an important parameter for the decision support of urban stormwater management. However, factors like comprehensive land-use type, variable spatial elevation, dynamic rainfall and groundwater elevation, make the direct estimation of runoff coefficient difficult. This paper presented a novel method to estimate the urban runoff coefficient using the inverse method, where observed time-series catchment outfall flow volume was employed as input for the water balance model and runoff coefficients of different catchments were treated as unknown parameters. A developed constrained minimization objective function was combined to solve the model and minimized error between observed and modeled outfall flow is satisfactory for the presenting of a set of runoff coefficients. Estimated runoff coefficients for the urban catchments in Shanghai downtown area demonstrated that practice of low impact design could play an important role in reducing the urban runoff.

Estimating emissions of HCFC-22 and CFC-11 in China by atmospheric observations and inverse modeling

This paper presents the estimation of Chinese emissions of HCFC-22 and CFC-11 in 2009 by an inverse modeling method based on in-situ measurement data from the Shangdianzi Global Atmosphere Watch (GAW) Regional Station (SDZ) and atmospheric transport simulations. After inversion (a-posteriori) estimates of the Chinese emissions in 2009 increased by 6.6% for HCFC-22 from 91.7 (± 83.6) to 98.3 (± 47.4) kt/yr and by 22.5% for CFC-11 from 13 (±12.6) to 15.8 (±7.2) kt/yr compared to an a-priori emission. While the model simulation with a-priori emissions already captured the main features of the observed variability at the measurement site, the model performance (in terms of correlation and mean-square-error) improved using a-posteriori emissions. The inversion reduced the root-mean-square (RMS) error by 4% and 10% for HCFC-22 and CFC-11, respectively.

Similarity measure of sedimentary successions and its application in inverse stratigraphic modeling

This paper presents a unique and formal method of quantifying the similarity or distance between sedimentary facies successions from measured sections in outcrop or drilled wells and demonstrates its first application in inverse stratigraphic modeling. A sedimentary facies succession is represented with a string of symbols, or facies codes in its natural vertical order, in which each symbol brings with it one attribute such as thickness for the facies. These strings are called attributed strings. A similarity measure is defined between the attributed strings based on a syntactic pattern-recognition technique. A dynamic programming algorithm is used to calculate the similarity. Inverse stratigraphic modeling aims to generate quantitative 3D facies models based on forward stratigraphic modeling that honors observed datasets. One of the key techniques in inverse stratigraphic modeling is how to quantify the similarity or distance between simulated and observed sedimentary facies successions at data locations in order for the forward model to condition the simulation results to the observed dataset such as measured sections or drilled wells. This quantification technique comparing sedimentary successions is demonstrated in the form of a cost function based on the defined distance in our inverse stratigraphic modeling implemented with forward modeling optimization.

Machine Learning-based Inverse Model for Few-Mode Fiber Designs

The medium for next-generation communication is considered as fiber for fast,secure communication and switching capability.Mode division and space division multiplexing provide an excellent switching capability with high data transmission rate.In this work,the authors have approached an inverse modeling technique using regression-based machine learning to design a weakly coupled few-mode fiber for facilitating mode division multiplexing.The technique is adapted to predict the accurate profile parameters for the proposed few-mode fiber to obtain the maximum number of modes.It is for a three-ring-core few-mode fiber for guiding five,ten,fifteen,and twenty modes.Three types of regression models namely ordinary least-square linear multi-output regression,k-nearest neighbors of multi-output regression,and ID3 algorithm-based decision trees for multi-output regression are used for predicting the multiple profile parameters.It is observed that the ID3-based decision tree for multioutput regression is the robust,highly-accurate machine learning model for fast modeling of FMFs.The proposed fiber claims to be an efficient candidate for the next-generation 5G and 6G backhaul networks using mode division multiplexing.

Retrieval of leaf biochemical properties by inversed PROSPECT model and hyperspectral indices:an application to Populus euphratica polymorphic leaves

Leaf biochemical properties have been widely assessed using hyperspectral reflectance information by inversion of PROSPECT model or by using hyperspectral indices, but few studies have focused on arid ecosystems. As a dominant species of riparian ecosystems in arid lands, Populus euphratica Oliv. is an unusual tree species with polymorphic leaves along the vertical profile of canopy corresponding to different growth stages. In this study, we evaluated both the inversed PROSPECT model and hyperspectral indices for estimating biochemical properties of P. euphratica leaves. Both the shapes and biochemical properties of P. euphratica leaves were found to change with the heights from ground surface. The results indicated that the model inversion calibrated for each leaf shape performed much better than the model calibrated for all leaf shapes, and also better than hyperspectral indices. Similar results were obtained for estimations of equivalent water thickness （EWT） and leaf mass per area （LMA）. Hyperspectral indices identified in this study for estimating these leaf properties had root mean square error （RMSE） and R2 values between those obtained with the two calibration strategies using the inversed PROSPECT model. Hence, the inversed PROSPECT model can be applied to estimate leaf biochemical properties in arid ecosystems, but the calibration to the model requires special attention.

Inverse Calculation of Wave-Absorbing Structure Dimensions Based on Extended ANFIS Model

A new wave energy dissipation structure is proposed, aiming to optimize the dimensions of the structure and make the reflection of the structure maintain a low level within the scope of the known frequency band. An optimal extended ANFIS model combined with the wave reflection coefficient analysis for the estimation of the structure dimensions is established. In the premise of lower wave reflection coefficient, the specific sizes of the structure are obtained inversely, and the contribution of each related parameter on the structural reflection performance is analyzed. The main influencing factors are determined. It is found that the optimal dimensions of the proposed structure exist, which make the wave absorbing performance of the structure reach a perfect status under a wide wave frequency band.

A case study of the consistency problem in the inverse estimation

Inverse technique is a widely used method in oceanography, but it has a problem that the retrieved solutions often violate model prior assumptions. To tune the model has consistent solutions, an iteration approach, which successively utilizes the posterior statistics for next round inverse estimation, is introduced and tested from a real case study. It is found that the consistency may become elusive as the determinants of solution and noise covariance matrices become zero in the iteration process. However, after several steps of such operation, the difference between posterior statistics and the model prior ones can be gradually reduced.

Inverse Learning Control of Nonlinear Systems Using Support Vector Machines

An inverse learning control scheme using the support vector machine (SVM) for regression was proposed. The inverse learning approach is originally researched in the neural networks. Compared with neural networks, SVMs overcome the problems of local minimum and curse of dimensionality. Additionally, the good generalization performance of SVMs increases the robustness of control system. The method of designing SVM inverse learning controller was presented. The proposed method is demonstrated on tracking problems and the performance is satisfactory.

Predictive Inverse Neurocontrol:an experimental case study

To increase predictive behaviors of neural network dynamic model, an experimental case study of a new approach to systems controller design is presented. The experiment is based on neural networks inverse plant model. Special rules for network training are developed. Such system is close to model-based predictive control, but needs much less computational resources. The approach advantages are shown by the control of laboratory complex plants.

Model predictive inverse method for recovering boundary conditions of two-dimensional ablation

A model predictive inverse method (MPIM) is presented to estimate the time- and space-dependent heat flux onthe ablated boundary and the ablation velocity of the two-dimensional ablation system. For the method, first of all, therelationship between the heat flux and the temperatures of the measurement points inside the ablation material is establishedby the predictive model based on an influence relationship matrix. Meanwhile, the estimation task is formulated as aninverse heat transfer problem (IHTP) with consideration of ablation, which is described by an objective function of thetemperatures at the measurement point. Then, the rolling optimization is used to solve the IHTP to online estimate theunknown heat flux on the ablated boundary. Furthermore, the movement law of the ablated boundary is reconstructedaccording to the estimation of the boundary heat flux. The effects of the temperature measurement errors, the numberof future time steps, and the arrangement of the measurement points on the estimation results are analyzed in numericalexperiments. On the basis of the numerical results, the effectiveness of the presented method is clarified.

Geochemical Modeling and Statistical Analysis for Groundwater Evolution Assessment in Wadi Qasab, Sohag, Eastern Desert, Egypt

Qasab basin is one of the most promising areas for the sustainable development in the Eastern Desert fringes of the Nile Valley, Egypt. The integration between statistical analysis, stable isotopes as well as geochemical modeling tools delineated the geochemical possesses affecting groundwater quality and detected the main recharge source in Qasab basin. The most of groundwater samples are brackish (88%), while the minority (12%) of the samples are fresh. The electrical conductivity of groundwater ranged from 1135 to 10,030 μS/cm. The statistical analysis and hydrochemical diagrams suggest that the groundwater quality is mainly controlled by several intermixed processes (rock weathering and agricultural activities). The mineralization of the Pleistocene groundwater is regulated by the rock weathering source, evaporation processes and reverse cation exchange. The isotopic signatures (δ2H and δ18O) represent two groundwater groups. The first group, is enriched with the isotopic signature of δ18O, which ranges from 0.9‰ to 5.5‰. This group is mostly affected by the recent meteoric recharge from the surface water leakage. The second group, is relatively depleted with the isotopic signature of δ18O, reflecting a palaeo recharge source of colder climate. The δ18O‰ varies from -10.1‰ to -6.4‰, indicating upward leakage of the Nubian sandstone aquifer through deep seated faults. The inverse geochemical model reflects that the salinity source of the groundwater samples is due to the leaching and dissolution processes of carbonate, sulphate and chloride minerals from the aquifer matrix. This study can demonstrate the hydrochemistry assessment guide to support sustainable development in Qasab basin to ensure that adequate groundwater management can play to reduce poverty and support socioeconomic development.

Improving the Accuracy of Vegetation Index Retrieval for Biomass by Combining Ground-UAV Hyperspectral Data-A New Method for Inner Mongolia Typical Grasslands

Grassland biomass is an important parameter of grassland ecosystems.The complexity of the grassland canopy vegetation spectrum makes the long-term assessment of grassland growth a challenge.Few studies have explored the original spectral information of typical grasslands in Inner Mongolia and examined the influence of spectral information on aboveground biomass(AGB)estimation.In order to improve the accuracy of vegetation index inversion of grassland AGB,this study combined ground and Unmanned Aerial Vehicle(UAV)remote sensing technology and screened sensitive bands through ground hyperspectral data transformation and correlation analysis.The narrow band vegetation indices were calculated,and ground and airborne hyperspectral inversion models were established.Finally,the accuracy of the model was verified.The results showed that:(1)The vegetation indices constructed based on the ASD FieldSpec 4 and the UAV were significantly correlated with the dry and fresh weight of AGB.(2)The comparison between measured R^(2) with the prediction R^(2) indicated that the accuracy of the model was the best when using the Soil-Adjusted Vegetation Index(SAVI)as the independent variable in the analysis of AGB(fresh weight/dry weight)and four narrow-band vegetation indices.The SAVI vegetation index showed better applicability for biomass monitoring in typical grassland areas of Inner Mongolia.(3)The obtained ground and airborne hyperspectral data with the optimal vegetation index suggested that the dry weight of AGB has the best fitting effect with airborne hyperspectral data,where y=17.962e^(4.672x),the fitting R^(2) was 0.542,the prediction R^(2)was 0.424,and RMSE and REE were 57.03 and 0.65,respectively.Therefore,established vegetation indices by screening sensitive bands through hyperspectral feature analysis can significantly improve the inversion accuracy of typical grassland biomass in Inner Mongolia.Compared with ground monitoring,airborne hyperspectral monitoring better reflects the inversion of actual surface biomass.It provides a reliable modeling framework for grassland AGB monitoring and scientific and technological support for grazing management.

A Hybrid Compensation Scheme for the Input Rate-Dependent Hysteresis of the Piezoelectric Ceramic Actuators

A hybrid compensation scheme for piezoelectric ceramic actuators(PEAs)is proposed.In the hybrid compensation scheme,the input rate-dependent hysteresis characteristics of the PEAs are compensated.The feedforward controller is a novel input rate-dependent neural network hysteresis inverse model,while the feedback controller is a proportion integration differentiation(PID)controller.In the proposed inverse model,an input ratedependent auxiliary inverse operator(RAIO)and output of the hysteresis construct the expanded input space(EIS)of the inverse model which transforms the hysteresis inverse with multi-valued mapping into single-valued mapping,and the wiping-out,rate-dependent and continuous properties of the RAIO are analyzed in theories.Based on the EIS method,a hysteresis neural network inverse model,namely the dynamic back propagation neural network(DBPNN)model,is established.Moreover,a hybrid compensation scheme for the PEAs is designed to compensate for the hysteresis.Finally,the proposed method,the conventional PID controller and the hybrid controller with the modified input rate-dependent Prandtl-Ishlinskii(MRPI)model are all applied in the experimental platform.Experimental results show that the proposed method has obvious superiorities in the performance of the system.

Direct shear tests on cemented paste backfill-rock wall and cemented paste backfill-backfill interfaces

Even though a large number of large-scale arch dams with height larger than 200 m have been built in the world, the transient groundwater flow behaviors and the seepage control effects in the dam foundations under difficult geological conditions are rarely reported. This paper presents a case study on the transient groundwater flow behaviors in the rock foundation of Jinping I double-curvature arch dam, the world＇s highest dam of this type to date that has been completed. Taking into account the geological settings at the site, an inverse modeling technique utilizing the time series measurements of both hydraulic head and discharge was adopted to back-calculate the permeability of the foundation rocks,which effectively improves the uniqueness and reliability of the inverse modeling results. The transient seepage flow in the dam foundation during the reservoir impounding was then modeled with a parabolic variational inequality（PVI） method. The distribution of pore water pressure, the amount of leakage, and the performance of the seepage control system in the dam foundation during the entire impounding process were finally illustrated with the numerical results.

A new method for multi-point pollution source identification

Rapid and accurate identification of the characteristics(source location,number,and intensity)of pollution sources is essential for emergency assessment of contamination events.Compared with single-point source iden-tification,the reconstruction of multiple sources is more challenging.In this study,a two-step inversion method is proposed for multi-point pollution source reconstruction from limited measurements with the number of sources unknown.The applicability of the proposed method is validated with a set of synthetic experiments correspond-ing to one-,two-,and three-point pollution sources.The results show that the number and locations of pollution sources are retrieved exactly the same as prescribed,and the source intensities are estimated with negligible errors.The algorithm exhibits good performance in single-and multi-point pollution source identification,and its accuracy and efficiency of identification do not deteriorate with the increase in the number of sources.Some limitations of the algorithm,together with its capabilities,are also discussed in this paper.

Geophysical and Production Data History Matching Based on Ensemble Smoother with Multiple Data Assimilation

The Ensemble Kalman Filter(EnKF),as the most popular sequential data assimilation algorithm for history matching,has the intrinsic problem of high computational cost and the potential inconsistency of state variables updated at each loop of data assimilation and its corresponding reservoir simulated result.This problem forbids the reservoir engineers to make the best use of the 4D seismic data,which provides valuable information about the fluid change inside the reservoir.Moreover,only matching the production data in the past is not enough to accurately forecast the future,and the development plan based on the false forecast is very likely to be suboptimal.To solve this problem,we developed a workflow for geophysical and production data history matching by modifying ensemble smoother with multiple data assimilation(ESMDA).In this work,we derived the mathematical expressions of ESMDA and discussed its scope of applications.The geophysical data we used is P-wave impedance,which is typically included in a basic seismic interpretation,and it directly reflects the saturation change in the reservoir.Full resolution of the seismic data is not necessary,we subsampled the P-wave impedance data to further reduce the computational cost.With our case studies on a benchmark synthetic reservoir model,we also showed the supremacy of matching both geophysical and production data,than the traditional reservoir history matching merely on the production data:the overall percentage error of the observed data is halved,and the variances of the updated forecasts are reduced by two orders of the magnitude.

Current studies on air distributions in commercial airliner cabins

Abstract Air distribution in commercial airliner cabins is very important for the comfort and health of passengers and crew. Experimental measurements, computational fluid dynamics （CFD） simulations, and inverse modeling are state-of-the-art methods available for studying the air distri- bution. This paper gave an overview of the different experimental models, such as scale models, simplified models, full-scale mockups, and actual air cabins. Although experimental measurements were expensive and time consuming, the data were essential for validating CFD simulations. Different modeling strategies for CFD simulations were also discussed in this paper, including large eddy simulations and Reynolds averaged Navier-Stokes equation modeling. CFD simulations were main stream approaches for studying the air distribution but they could not easily lead to optimal design. Inverse modeling of air distribution has recently emerged into a very powerful and attractive tool for designing the air distribution in airliner cabins, although most of the studies were preliminary.

Genetic Algorithm-Based Estimation of Nonlinear Transducer

This paper describes an innovative, genetic algorithm based inverse model of nonlinear transducer. In the inverse modeling, using a genetic algorithm, the unknown coefficients of the model are estimated accurately. The simulation results indicate that this technique provides greater flexibility and suitability than the existing methods. It is very easy to modify the nonlinear transducer on line. Thus the method improves the transducer's accuracy. With the help of genetic algorithm (GA), the model coefficients' training are less likely to be trapped in local minima than traditional gradient based search algorithms.