Error Elimination Algorithm Research Articles

TMR sensor current measurement technology for overhead lines

In order to meet the needs of TMR current sensors for overhead line high precision current measurement, a neighboring conductor interference magnetic field error elimination algorithm is proposed. Based on the Biot-Savard law, through the establishment of an interference magnetic field environment mathematical model, the accuracy of the error elimination algorithm is verified in maxwell simulation software. The simulation results show that under the working frequency current, the adjacent wire interference magnetic field elimination algorithm can reduce the error from 14% to less than 1%, and under the lightning wave environment, the error is reduced from 19% to less than 1%. This proves that the proposed algorithm is effective in reducing the measurement error caused by the interference magnetic field of the adjacent conductor.

Non-invasive biosignals detection for continuous monitoring of a neonate using quartz crystal resonator

A wide-range load sensing device and signal processing algorithm were presented for the non-invasive monitoring of a neonate. The device was developed based on a quartz crystal resonator (QCR) that exhibited wide measurement range to the applied load. The resolution of the force sensor was improved to detect the small force generated by heartbeats and respirations of a neonate. The force signals of a neonate laid on a bed were successfully measured in the craniocaudal direction by using the device installed under the bed. In addition, an error elimination algorithm was developed based on the Fast Fourier Transform (FFT) analysis to detect the biosignals accurately from the load data. Using the proposed method, we achieved the detection of the heart rate with the average error of 0.7 beats per minute (bpm) and standard deviation of 3.5 bpm and respiratory rate with the average error of −1.6 breaths per minute (brpm) and standard deviation of 7.1 brpm respectively.

Application of 3D laser scanning technology for image data processing in the protection of ancient building sites through deep learning

To study the conservation of architectural heritage, computer algorithms are used to process the artificial field measurements and reference data maps. A 3D virtual model of traditional architecture is constructed for the direct learning and feature extraction from image data through deep learning. A set of tools and processes for pixel-level image processing labeling are proposed, and a label quality checking tool is written specially. Through the training method of transfer learning, the convergence speed and accuracy of the network are accelerated and improved. Based on the field data collection of the architectural heritage by a 3D laser scanner, the impacts of the number of stations, the number of targets, and the distance on the data scanning results are analyzed, thereby drawing the rules and principles of setting up the survey stations. While processing the point cloud data, for the redundant data and the rough difference points found in the original point cloud data, the program is written by the gross error elimination algorithm, which realizes the automatic elimination of the point cloud gross error data and provides a convenient method for data processing. The data collection, data processing, and model construction of architectural heritage are performed by the 3D laser scanner. The 3D models of traditional architecture with texture photos are obtained. The results show that through correlation error analysis and model evaluation, the requirements for the measurement of traditional architecture can be achieved. Therefore, the technology has a guiding significance for the conservation of traditional architecture, which proves that the 3D laser scanner has broad application prospects in the surveying and mapping of architectural heritage.

Static Current Error Elimination Algorithm for Induction Motor Predictive Current Control

Predictive current control of induction motors can effectively avoid performance deterioration of control caused by delays in the current loop and improve the dynamic performance of current control. However, owing to measurement errors and parameter changes, deviations can appear between the predictive controller parameters and the actual motor parameters. This might lead to static current error, which can cause problems, including decrease in system’s efficiency, inability to deliver nominal torque, and to operate in torque control mode, among others. Based on an induction motor model, this paper quantitatively analyzes the influence on current control stability caused by errors in the predictive control model parameters. In addition, we present the mathematical relation between errors in model parameters and static current error, and propose an algorithm to eliminate this type of error. The algorithm corrected the parameters for predictive control using $dq$ axis current feedback and eliminated the static error caused by parameter mismatch. Through experimental results, the stability and effectiveness of the proposed method were shown.

Improved Indoor Location Systems in a Controlled Environments

The precise localization by using Wi-Fi Access Point (AP) has become a very important issue for indoor location based services such as marketing, patient follow up and so on. Present AP localization systems are working on specially designed Wi-Fi units, and their algorithms using radio signal strength (RSS) exhibit (relatively) high errors, so industry looks more precise and fast adaptable methods. A new model considering/eliminating strong RSS levels in addition to close distance error elimination algorithm (CDEEA) combined with median filters has been proposed in order to increase the performance of conventional RSS based location systems. Collecting local signal strengths by means of an ordinary WiFi units present on any laptop as a receiver is followed by the application of CDEEA to eliminate strong RSS levels. Median filter is then applied to those eliminated values, and AP based path loss model is generated, adaptivelly. Finally, the proposed algorithm predicts locations within a maximum mean error of 2.96m for 90% precision level. This achievement with an ordinary wifi units present on any commercial laptop is comparably at very good level in literature.

Orbit determination and time synchronization for a GEO/IGSO satellite navigation constellation with regional tracking network

Aiming at regional services, the space segment of COMPASS (Phase I) satellite navigation system is a constellation of Geostationary Earth Orbit (GEO), Inclined Geostationary Earth Orbit (IGSO) and Medium Earth Orbit (MEO) satellites. Precise orbit determination (POD) for the satellites is limited by the geographic distribution of regional tracking stations. Independent time synchronization (TS) system is developed to supplement the regional tracking network, and satellite clock errors and orbit data may be obtained by simultaneously processing both tracking data and TS data. Consequently, inconsistency between tracking system and TS system caused by remaining instrumental errors not calibrated may decrease navigation accuracy. On the other hand, POD for the mixed constellation of GEO/IGSO/MEO with the regional tracking network leads to parameter estimations that are highly correlated. Notorious example of correlation is found between GEO’s orbital elements and its clock errors. We estimate orbital elements and clock errors for a 3GEO+2IGSO constellation in this study using a multi-satellite precise orbit determination (MPOD) strategy, with which clock error elimination algorithm is applied to separate orbital and clock estimates to improve numerical efficiency. Satellite Laser Ranging (SLR) data are used to evaluate User Ranging Error (URE), which is the orbital error projected on a receiver’s line-of-sight direction. Two-way radio-wave time transfer measurements are used to evaluate clock errors. Experimenting with data from the regional tracking network, we conclude that the fitting of code data is better than 1 m in terms of Root-Mean-Square (RMS), and fitting of carrier phase is better than 1 cm. For orbital evaluation, difference between computed receiver-satellite ranging based on estimated orbits and SLR measurements is better than 1 m (RMS). For clock estimates evaluation, 2-hour linear-fitting shows that the satellite clock rates are about 1.E-10 s/s, while receiver clock rates are about 1×10−13–1×10−12 s/s. For the 72-hour POD experiment, the average differences between POD satellite clock rates estimates and clock measurements based on TS system are about 1×10−13 s/s, and for receiver clock rates, the differences are about 1×10−15 s/s.

An analytical driver model for arbitrary path following at varying vehicle speed

In this paper, an analytical driver model in the form of a simple artificial neural network is presented. Based on the preview-follower theory and error elimination algorithm, all the driver model parameters can be calculated analytically. In order to extend the driver model's application scope, all the driver parameters are fitted as the functions of vehicle speed, and a preview point searching algorithm is proposed to enable the driver model to follow an arbitrary designated path defined in ground coordinate. Driver/vehicle closed-loop simulations indicate that the proposed driver model is suitable for arbitrary path following at varying vehicle speed.

An error elimination algorithm of a pole-assignment self-tuner for step reference inputs

This paper proposes a modified version of pole assignment self-tuners for linear time-invariant systems subjected to unknown constant disturbances. It is shown that this version also has the local convergence property established by Goodwin and Sin(1981, 1984) which was developed for linear systems with no disturbances. Utilizing this property, an error elimination algorithm is developed for step references. The performance of this algorithm is demonstrated through a simulation and the result is compared with that of a conventional pole assignment self-tuner using an additional integral action.