Off-line Estimation Research Articles

Constrained Least-Squares Parameter Estimation for a Double Layer Capacitor

This paper presents an estimation of the parameters for a Double Layer Super Capacitor (DLC) that is modelled with a two-branch circuit. The estimation is achieved using a constrained minimization technique, which is developed off-line and uses a single constraint to write the matrix equation. The model is algebraically manipulated to obtain a matrix equation, and a signal processing system is developed to prepare the signals for the identification algorithms. The proposed method builds on the results obtained using an unconstrained ordinary least-squares (OLS) technique. The method is tested both in simulation and experimentally, using a specially-designed experimental rig. A current ramp input is used to generate the corresponding output voltage and its derivatives. The results obtained from the constrained off-line minimization algorithm are compared with those obtained using a traditional off-line estimation method. The discussion of the results shows that the proposed method outperforms the traditional estimation technique. In summary, this paper contributes to the field of DLC parameter estimation by introducing a new off-line constrained minimization technique. The results obtained from the simulations and experimental rig demonstrate the effectiveness of the proposed method with two of three parameters showing relative errors less than 5%.

A high‐precision and fast off‐line beamforming pattern estimation technique for active phased array antenna application

A detailed modeling method including all states of each active devices is proposed in this work, aiming to achieve high-precision and fast off-line estimation of beamforming pattern. Experimental verification of hundreds of radiating patterns shows that the beam steering angle accuracy is able to obtain 0.2% (Tx mode) and 0.19% (Rx mode) beam width in average, respectively; and the beam width accuracy is 0.2% (Tx mode) and 0.39% (Rx mode) in average, respectively. Also, SLL accuracy can reach 0.01 dB (Tx mode) and 0.1 dB (Rx mode), respectively. By adopting this technology, the optimal beam controlling codes of large active phased array antenna (APAA) could be quickly derived. Furthermore, measurement and online compensation works of enormous quantities of beams could be greatly simplified.

State-of-health estimation based on real data of electric vehicles concerning user behavior

State of health (SOH) of lithium-ion battery pack directly determines the driving mileage and output power of the electric vehicle. With the development of big data storage and analysis technology, using big data to off-line estimate battery pack SOH is more feasible than before. This paper proposes a SOH estimation method based on real data of electric vehicles concerning user behavior. The charging capacity is calculated by historical charging data, and locally weighted linear regression (LWLR) algorithm is used to qualitatively characterize the capacity decline trend. The health features are extracted from historical operating data, maximal information coefficient (MIC) algorithm is used to measure the correlation between health features and capacity. Then, long and short-term memory (LSTM)-based neural network will further learn the nonlinear degradation relationship between capacity and health features. Bayesian optimization algorithm is used to ensure the generalization of the model when different electric vehicles produce different user behaviors. The estimation method is validated by the 300 days historical dataset from 100 vehicles with different driving behavior. The results indicates that the maximum relative error of estimating SOH is 0.2%.

Experimental investigation on the potential of the Feedback Fx-LMS Active Noise Control technology for use in industrial generator sets

Active Noise Control (ANC) has been considered a promising technology for the abatement of acoustic noise from the mid-20th century. Feedback and Feedforward ANC algorithms, based on the destructive interference principle applied to acoustic waves, have been developed for different applications, depending on the spectrum of the noise source. Feedback ANC algorithms make use of a single control microphone to measure an error signal which is then employed by an adaptive filter to estimate the noise source and generate an opposite-phase control signal. The Fx-LMS (Filtered-X Least Mean Square) algorithm is mostly adopted to update the filter. Feedback ANC systems have proven to be effective for the abatement of low-frequency quasi-steady noises; however, different challenges must be overcome to realize an effective and durable system for high-temperature application. This paper aims at experimentally assessing the feasibility of a Feedback Fx-LMS ANC system with off-line Secondary Path estimation to be used in mid-size diesel gensets for the reduction of the exhaust noise. Several solutions are proposed, including the mechanical design, the development of the Fx-LMS algorithm in the LabVIEW FPGA programming language, and the key features required to prevent parts from thermal damage and fouling. The developed prototype was implemented on a 50-kW diesel genset and tested in a semi-anechoic chamber. The noise abatement inside the exhaust pipe and at different measurement points around the machine was evaluated and discussed, showing good potential for improving the acoustic comfort of genset users.

Occupancy Estimation in Buildings Based on Infrared Array Sensors Detection

Along with methods such as Hidden Markov Model and Linear Discriminant Analysis, environmental sensors including temperature, humidity or carbon dioxide are used to estimate indoor occupancy, which have various applications. Most of the previous studies have neglected the time-dependent character of the indoor occupancy information or real-time response of the system. In this paper, we propose to use low invasive, fast-sampling infrared array sensors to collect data from the actual scene and establish the Inhomogeneous Hidden Markov Model to capture the time dependence of occupancy for buildings occupancy estimation. First, to avoid raw sensor datas susceptibility to external influences, the non-negative matrix factorization is adopted to reduce the dimension of the raw matrix and eliminate interferences caused by environmental changes. Second, the Softmax Regression Model is used to calculate the emission probability matrix for clarifying the dynamic relationship between environmental parameters and indoor occupancy, which is weak. Third, the Forward algorithm and the Viterbi algorithm are applied to achieve online and off-line estimation. Experiments are made with real recorded data. Our performance results demonstrate that method we proposed is effective for indoor occupancy estimation.

Walking Control of a Biped Robot on Static and Rotating Platforms Based on Hybrid Reinforcement Learning

In this paper, we proposed a novel Hybrid Reinforcement Learning framework to maintain the stability of a biped robot (NAO) while it is walking on static and dynamic platforms. The reinforcement learning framework consists of the Model-based off-line Estimator, the Actor Network Pre-training scheme, and the Mode-free on-line optimizer. We proposed the Hierarchical Gaussian Processes as the Mode-based Estimator to predict a rough model of the system and to obtain the initial control input. Then, the initial control input is employed to pre-train the Actor Network by using the initial control input. Finally, a model-free optimizer based on Deep Deterministic Policy Gradient framework is introduced to fine tune the Actor Network and to generate the best actions. The proposed reinforcement learning framework not only successfully avoids the distribution mismatch problem while combining model-based scheme with model-free structure, but also improves the sample efficiency for the on-line learning procedure. Simulation results show that the proposed Hybrid Reinforcement Learning mechanism enables the NAO robot to maintain balance while walking on static and dynamic platforms. The robustness of the learned controllers in adapting to platforms with different angles, different magnitudes, and different frequencies is tested.

Depolarizing Channel Mismatch and Estimation Protocols for Quantum Turbo Codes

Quantum turbo codes (QTC) have shown excellent error correction capabilities in the setting of quantum communication, achieving a performance less than 1 dB away from their corresponding hashing bounds. Decoding for QTCs typically assumes that perfect knowledge about the channel is available at the decoder. However, in realistic systems, such information must be estimated, and thus, there exists a mismatch between the true channel information and the estimated one. In this article, we first heuristically study the sensitivity of QTCs to such mismatch. Then, existing estimation protocols for the depolarizing channel are presented and applied in an off-line manner to provide bounds on how the use of off-line estimation techniques affects the error correction capabilities of QTCs. Finally, we present an on-line estimation method for the depolarizing probability, which, different from off-line estimation techniques, neither requires extra qubits, nor increases the latency. The application of the proposed method results in a performance similar to that obtained with QTCs using perfect channel information, while requiring less stringent conditions on the variability of the channel than off-line estimation techniques.

Identification of continuous-time models with slowly time-varying parameters

The off-line estimation of the parameters of continuous-time, linear, time-invariant transfer function models can be achieved straightforwardly using linear prefilters on the measured input and output of the system. The on-line estimation of continuous-time models with time-varying parameters is less straightforward because it requires the updating of the continuous-time prefilter parameters. This paper shows how such on-line estimation is possible by using recursive instrumental variable approaches. The proposed methods are presented in detail and also evaluated on a numerical example using both single experiment and Monte Carlo simulation analysis. In addition, the proposed recursive algorithms are tested using data from two real-life systems.

Cardiac output estimation using multi-beat analysis of the radial arterial blood pressure waveform: a method comparison study in patients having off-pump coronary artery bypass surgery using intermittent pulmonary artery thermodilution as the reference method

Pulse wave analysis enables stroke volume to be estimated from an arterial blood pressure waveform. Multi-beat analysis is a novel pulse wave analysis method. We aimed to investigate cardiac output (CO) estimations using multi-beat analysis of the radial arterial blood pressure waveform in patients undergoing off-pump coronary artery bypass surgery (OPCAB) using intermittent pulmonary artery thermodilution (PATD) as the reference method. This was a prospective clinical method comparison study. In 58 patients, we measured CO using PATD (PATD-CO; reference method) and simultaneously recorded the radial arterial blood pressure waveform that we used for off-line estimation of CO based on multi-beat analysis (MBA-CO; test method) using the Argos CO monitor (Retia Medical; Valhalla, NY, USA). The final analysis was performed using 572 paired CO measurements. We performed Bland–Altman analysis accounting for multiple observations per patient. To describe the ability of the test method to track changes in CO over time we computed four-quadrant plots using a central exclusion zone of 15% and calculated the concordance rate. Mean PATD-CO was 4.13 ± 1.26 L/min and mean MBA-CO was 4.31 ± 1.25 L/min. The mean of the differences between PATD-CO and MBA-CO was − 0.20 L/min with a standard deviation of ± 1.14 L/min and 95% limits of agreement of − 2.48 to + 2.08 L/min. The concordance rate for CO changes between PATD-CO and MBA-CO was 89%. CO estimations using multi-beat analysis (Argos monitor) show reasonable agreement and trending ability compared with PATD-CO as the reference method in adult patients during OPCAB.

Testing the limits of virtual deep seismic sounding via new crustal thickness estimates of the Australian continent

We apply virtual deep seismic sounding (VDSS) to data collected from both permanent and temporary seismic stations in Australia with the goal of examining (i) the resilience of the method to the presence of complex lithospheric structure, and (ii) the effectiveness of different approaches for estimating bulk crustal properties (namely thickness and Vp). Data from the permanent WRAB in the Northern Territory station is ideal for benchmarking VDSS (large number and favourable distribution of recorded earthquakes), with the results from several approaches agreeing on a thickness of 40-42 km. Application of VDSS to data from the temporary BILBY array, a linear distribution of broadband stations that traverses central Australia, shows that strong Moho reflections can be retrieved with as few as two earthquakes even at the transition between crustal blocks of different character and in the presence of thick sedimentary basins. Crustal thickness varies between 36-54 km and compare well with the reflectivity character of nearby deep seismic reflection lines. Furthermore, we find that off-line estimates of crustal thickness, calculated by binning the source regions according to back-azimuth, produce estimates of crustal thickness that are consistent with the regional geology. Overall, we find that VDSS is a powerful technique for estimating crustal thickness and velocity due to its insensitivity to complex short-wavelength structure and requirement of a small number earthquakes to produce a stable result. However, not all schemes tested for extracting bulk crustal properties appear to be robust and stringent data quality checking is still required during implementation.

Estimation of hydrodynamic parameters for underwater systems using a simple off-line regression method: a case study

Estimation of hydrodynamic model parameters such as added mass parameter and drag coefficients is very crucial while mathematically modelling any underwater system. Owing to the nature of the model being coupled and its high non-linearity, estimation of these parameters becomes complicated. Various methods of parameter estimation have been currently employed, involving experiments, computational fluid analysis, and simulations, each having its own advantages and disadvantages. Most of the existing methods use sophisticated external instrumentation and experimentally estimate the parameters using computationally expensive adaptive algorithms, which may not be required or cannot be generalised to all the systems. In this paper, a simple off-line estimation method for estimating crucial hydrodynamic parameters using onboard sensors is presented. The error between the data from numerical simulations (using a mathematical model) and experiments is iteratively minimised using a gradient descent-based optimisation algorithm, by having the unknown model parameters as design variables of the error minimisation process. The method combines the properties of least squares estimation and the free decay tests, where the system can be excited with any known external inputs. A mathematical model with the unknown parameters, fully defining the behaviour of the system, is required and open loop experimental data from onboard sensors for a known input is sufficient for the estimation process, thereby eliminating the requirement of additional instrumentation. Non-linear mathematical models can be directly used in estimation, unlike few other methods which require linearisation and approximation. This method can be generalised to any system, provided sufficient information on experimental input and output, and the equivalent mathematical model of the system are available. The proposed method has been successfully implemented to estimate the added mass and drag coefficients of a standalone, single degree of freedom variable buoyancy module, ‘vBuoy’. A mathematical model defining the dynamics of the heave motion of vBuoy has been derived and the parameters involved in the model have been estimated with the proposed method. The proposed method, as well as the results, are validated by comparing the experiments and simulations at different conditions. The results showed that the proposed method was well suited for the estimation of hydrodynamic parameters of underwater systems.

An instrumental variable method for robot identification based on time variable parameter estimation

This paper considers the data-based identification of industrial robots using an instrumental variable method that uses off-line estimation of the joint velocities and acceleration signals based only on the measurement of the joint positions. The usual approach to this problem relies on a ‘tailor-made’ prefiltering procedure for estimating the derivatives that depends on good prior knowledge of the system’s bandwidth. The paper describes an alternative Integrated Random Walk SMoothing (IRWSM) method that is more robust to deficiencies in such a priori knowledge and exploits an optimal recursive algorithm based on a simple integrated random walk model and a Kalman filter with associated fixed interval smoothing. The resultant IDIM-IV instrumental variable method, using this approach to signal generation, is evaluated by its application to an industrial robot arm and comparison with previously proposed methods.

Analysis of Correlation in MEMS Gyroscope Array and its Influence on Accuracy Improvement for the Combined Angular Rate Signal.

Obtaining a correlation factor is a prerequisite for fusing multiple outputs of a mircoelectromechanical system (MEMS) gyroscope array and evaluating accuracy improvement. In this paper, a mathematical statistics method is established to analyze and obtain the practical correlation factor of a MEMS gyroscope array, which solves the problem of determining the Kalman filter (KF) covariance matrix Q and fusing the multiple gyroscope signals. The working principle and mathematical model of the sensor array fusion is briefly described, and then an optimal estimate of input rate signal is achieved by using of a steady-state KF gain in an off-line estimation approach. Both theoretical analysis and simulation show that the negative correlation factor has a favorable influence on accuracy improvement. Additionally, a four-gyro array system composed of four discrete individual gyroscopes was developed to test the correlation factor and its influence on KF accuracy improvement. The result showed that correlation factors have both positive and negative values; in particular, there exist differences for correlation factor between the different units in the array. The test results also indicated that the Angular Random Walk (ARW) of 1.57°/h0.5 and bias drift of 224.2°/h for a single gyroscope were reduced to 0.33°/h0.5 and 47.8°/h with some negative correlation factors existing in the gyroscope array, making a noise reduction factor of about 4.7, which is higher than that of a uncorrelated four-gyro array. The overall accuracy of the combined angular rate signal can be further improved if the negative correlation factors in the gyroscope array become larger.

An off-line output feedback MPC strategy for nonlinear systems represented by quasi-LPV model

In case when a nonlinear system is represented by quasi-LPV model with bounded disturbance, we adopt the parameter-dependent dynamic output feedback MPC (PDDOFMPC) with guaranteed quadratic boundedness and physical constraints. We pre-specify one sequence of nested ellipsoids for estimated state, and another for estimation error, then calculate PDDOFMPC parameters for each combination of estimated state ellipsoid and estimation error ellipsoid. A look-up table is constructed off-line to store these controller parameters each related to a unique combination. On-line, the estimated state is iterated and the estimation error set is refreshed. By checking the smallest off-line estimated state ellipsoid to contain the real-time estimated state, and the smallest off-line estimation error ellipsoid to include the real-time estimation error set, a unique set of control parameters is taken from the look-up table at each sampling instant. An example is given to illustrate the effectiveness of the approach.

Self-weighted recursive estimation of GARCH models

ABSTRACTThe generalized autoregressive conditional heteroscedasticity (GARCH) processes are frequently used to investigate and model financial returns. They are routinely estimated by computationally complex off-line estimation methods, for example, by the conditional maximum likelihood procedure. However, in many empirical applications (especially in the context of high-frequency financial data), it seems necessary to apply numerically more effective techniques to calibrate and monitor such models. The aims of this contribution are: (i) to review the previously introduced recursive estimation algorithms and to derive self-weighted alternatives applying general recursive identification instruments, and (ii) to examine these methods by means of simulations and an empirical application.

Simultaneous Recognition and Relative Pose Estimation of 3D Objects Using 4D Orthonormal Moments.

Both three-dimensional (3D) object recognition and pose estimation are open topics in the research community. These tasks are required for a wide range of applications, sometimes separately, sometimes concurrently. Many different algorithms have been presented in the literature to solve these problems separately, and some to solve them jointly. In this paper, an algorithm to solve them simultaneously is introduced. It is based on the definition of a four-dimensional (4D) tensor that gathers and organizes the projections of a 3D object from different points of view. This 4D tensor is then represented by a set of 4D orthonormal moments. Once these moments are arranged in a matrix that can be computed off-line, recognition and pose estimation is reduced to the solution of a linear least squares problem, involving that matrix and the 2D moments of the observed projection of an unknown object. The abilities of this method for 3D object recognition and pose estimation is analytically proved, demonstrating that it does not rely on experimental work to apply a generic technique to these problems. An additional strength of the algorithm is that the required projection is textureless and defined at a very low resolution. This method is computationally simple and shows very good performance in both tasks, allowing its use in applications where real-time constraints have to be fulfilled. Three different kinds of experiments have been conducted in order to perform a thorough validation of the proposed approach: recognition and pose estimation under z axis (yaw) rotations, the same estimation but with the addition of y axis rotations (pitch), and estimation of the pose of objects in real images downloaded from the Internet. In all these cases, results are encouraging, at a similar level to those of state-of-the art algorithms.

P371 Simulation and estimation of Hawkes processes for modeling EEG data to forecast epileptic seizures

Objectives Prediction of epileptic seizures has been in the forefront of the research for a long time but without the development of a satisfactory solution so far. Our aim is to develop a forecasting method – previously applied in seismology – that is based on fitting so-called Hawkes processes to EEG data, both exhibiting self-exiting dynamics. The choice of this approach is justified by the observation of Sornette and Osorio, according to which earthquakes and epileptic seizures show numerous similarities in their dynamics. Methods As an essential preliminary work for model-fitting we first implemented an algorithm for the simulation of self-exiting point processes. This was followed by the implementation of a maximum likelihood estimation (MLE) method based on the work of Ozaki. Results By visual inspection of projections of the three-dimensional parameter space we found a variety of behaviors of the cost function including some degenerate constellations of the parameters, where the behavior of the cost function relies almost solely on one parameter. Despite the simplicity of the local optimization method applied during the off-line estimation of the parameters, we succeeded to estimate the parameters with a fairly good approximation. Discussion and conclusion Based on our preliminary results, we conclude that the simulation of Hawkes-processes and fitting them to a point-process derived from EEG data is promising. As a next step, we plan to implement an on-line MLE method, by which real-time change detection in the EEG-derived parameters becomes possible. Significance As far as we know, self-exiting point processes have not been applied in seizure prediction before.

Estimation of UAV Position with Use of Smoothing Algorithms

Abstract The paper presents methods of on-line and off-line estimation of UAV position on the basis of measurements from its integrated navigation system. The navigation system installed on board UAV contains an INS and a GNSS receiver. The UAV position, as well as its velocity and orientation are estimated with the use of smoothing algorithms. For off-line estimation, a fixed-interval smoothing algorithm has been applied. On-line estimation has been accomplished with the use of a fixed-lag smoothing algorithm. The paper includes chosen results of simulations demonstrating improvements of accuracy of UAV position estimation with the use of smoothing algorithms in comparison with the use of a Kalman filter.

Dynamic estimation of the modeling error statistics in Diffuse Optical Tomography

Diffuse Optical Tomography (DOT) is a non-invasive imaging technique that suffers from a typical large-scale and ill-posed inverse problem with low spatial resolution. In DOT, the inverse problem is computationally intensive and decreasing the computation complexity and making it well-posed is the one of the most challenging research areas. More precisely, one of the well-known complexity reduction techniques is defined as applying modelling error originated from discretization of forward problem. Applying the discretization error in Bayesian inference has already been discussed; the method in which the likelihood is modified by an off-line prior density estimation. This paper implements a new method to enhance the modelling error approach using an iterative scheme to update statistical parameters of modelling discrepancy in DOT. The algorithm is very similar to Ensemble Kalman Filter. Moreover, the reconstruction process in the applied method is conducted by a small sample size rather than off-line method. Hence, the computation complexity is decreased and the algorithm converges in few iterations. The efficiency of the proposed method is illustrated by simulations.

Development of Macroscopic Emission Estimation Model Based on Microscopic Operating Modes

This study derived two models to estimate emissions on the basis of macroscopic mobility measures. The developed models can be used as part of signal timing optimization, macroscopic and mesoscopic simulation, cost-benefit analysis, and real-time and off-line estimation of arterial road performance. The derivation relied on output from the Motor Vehicle Emission Simulator model operating mode distribution analysis and used estimates from microscopic simulation model runs as input. Models with various combinations of independent variables were tested and two final models were selected. The first model used vehicle miles traveled (VMT) and vehicle delays as independent variables. The second model used VMT, vehicle delays, stop delays, and number of stops. Both models showed acceptable performance for most of the investigated cases in simulation modeling. A further test was performed using vehicle trajectories from the next generation simulation program databases. The results showed that the model developed on the basis of VMT and total vehicle delays performed well. However, the second model, developed on the basis of VMT, total vehicle delays, total stop delays, and number of stops, did not perform well. This could be because the model required accurate vehicle trajectories, including accurate acceleration and deceleration estimates. The quality of the trajectories in microscopic simulation is expected to affect the accuracy of the second model.