Diagnosis of COVID-19 Pneumonia Based on Graph Convolutional Network.

A three-dimensional (3D) deep learning method is proposed, which enables the rapid diagnosis of coronavirus disease 2019 (COVID-19) and thus significantly reduces the burden on radiologists and physicians. Inspired by the fact that the current chest computed tomography (CT) datasets are diversified in equipment types, we propose a COVID-19 graph in a graph convolutional network (GCN) to incorporate multiple datasets that differentiate the COVID-19 infected cases from normal controls. Specifically, we first apply a 3D convolutional neural network (3D-CNN) to extract image features from the initial 3D-CT images. In this part, a transfer learning method is proposed to improve the performance, which uses the task of predicting equipment type to initialize the parameters of the 3D-CNN structure. Second, we design a COVID-19 graph in GCN based on the extracted features. The graph divides all samples into several clusters, and samples with the same equipment type compose a cluster. Then we establish edge connections between samples in the same cluster. To compute accurate edge weights, we propose to combine the correlation distance of the extracted features and the score differences of subjects from the 3D-CNN structure. Lastly, by inputting the COVID-19 graph into GCN, we obtain the final diagnosis results. In experiments, the dataset contains 399 COVID-19 infected cases, and 400 normal controls from six equipment types. Experimental results show that the accuracy, sensitivity, and specificity of our method reach 98.5%, 99.9%, and 97%, respectively.

An Ultra-Short Baseline Positioning Model Based on Rotating Array & Reusing Elements and Its Error Analysis.

The USBL (Ultra-Short Base Line) positioning system is widely used in underwater acoustic positioning systems due to its small size and ease of use. The traditional USBL positioning system is based on 'slant range and azimuth'. The positioning error is an increasing function with the increase in distance and the positioning accuracy depends on the ranging accuracy of the underwater target. This method is not suitable for long-distance underwater positioning operations. This paper proposes a USBL positioning calculation model based on depth information for 'rotating array and reusing elements'. This method does not need to measure the distance between the USBL acoustic array and target, so it can completely eliminate the influence of long-distance ranging errors in USBL positioning. The theoretical analysis and simulation experiments show that the new USBL positioning model based on 'rotating array and reusing elements' can completely eliminate the influence of the wavelength error and spacing error of underwater acoustic signals on the positioning accuracy of USBL. The positioning accuracy can be improved by approximately 90%, and the horizontal positioning error within a positioning distance of 1000 m is less than 1.2 m. The positioning method has high precision performance in the long distance, and provides a new idea for the engineering design of a USBL underwater positioning system.

A misalignment angle error calibration method of underwater acoustic array in strapdown inertial navigation system/ultrashort baseline integrated navigation system based on single transponder mode.

The angle misalignment error of a USBL (Ultrashort Baseline) acoustic array is one of the major error sources of the strapdown inertial navigation system/USBL positioning system, which will directly affect the positioning accuracy of the USBL positioning system. For the traditional calibration method cannot accurately estimate the angle misalignment error due to its strict trajectory requirements in the field experiment and the high-precision layout of the transceiver array elements, a new method for estimating the angle misalignment error of a USBL acoustic array based on single transponder and dual-vector reconstruction is studied in this paper. The precondition of USBL misalignment calibration is to locate the underwater transponder accurately. In this paper, the single transponder segmentation iterative long baseline method is used to locate the underwater target transponder. The dual-vector reconstruction method is studied to control the estimation accuracy of USBL misalignment error calibration based on the traditional single transponder method, which provides a theoretical basis for the determination of the iteration times to the USBL angle misalignment error estimation module. The underwater experiment results show that the positioning error could be reduced to less than 1 m after the angle misalignment error compensation. The underwater transponder positioning and the angle misalignment error estimation of USBL could be accomplished in a circle sailing. It is a new method with good performance of high estimation accuracy, simple operation, and easy realization.

A calibration method of ultra-short baseline installation error with large misalignment based on variational Bayesian unscented Kalman filter.

The ultrashort baseline system is widely used in ships and underwater navigation and positioning. The misalignment and level arm with the inertial measurement unit are the two sources of positioning inaccuracy. The accuracy of calibration is usually affected by measurement noise and linearization of the observation equation. In order to improve the calibration accuracy, the variational Bayesian unscented Kalman filter (VBUKF) method is proposed for the calibration of the ultrashort baseline installation error in the paper. The detailed derivation of VBUKF for the calibration is presented in the paper. Simulation experiments and field experiments were carried out, respectively, to verify the algorithm. The simulation results show that the proposed method can calibrate the installation error of the sensors in real time on line. The field experiment verified that the algorithm improves the calibration accuracy of the installation error under the large misalignment. The positioning accuracy is also improved compared with the traditional method.

Coarse Alignment Technology on Moving base for SINS Based on the Improved Quaternion Filter Algorithm.

Initial alignment of the strapdown inertial navigation system (SINS) is intended to determine the initial attitude matrix in a short time with certain accuracy. The alignment accuracy of the quaternion filter algorithm is remarkable, but the convergence rate is slow. To solve this problem, this paper proposes an improved quaternion filter algorithm for faster initial alignment based on the error model of the quaternion filter algorithm. The improved quaternion filter algorithm constructs the K matrix based on the principle of optimal quaternion algorithm, and rebuilds the measurement model by containing acceleration and velocity errors to make the convergence rate faster. A doppler velocity log (DVL) provides the reference velocity for the improved quaternion filter alignment algorithm. In order to demonstrate the performance of the improved quaternion filter algorithm in the field, a turntable experiment and a vehicle test are carried out. The results of the experiments show that the convergence rate of the proposed improved quaternion filter is faster than that of the tradition quaternion filter algorithm. In addition, the improved quaternion filter algorithm also demonstrates advantages in terms of correctness, effectiveness, and practicability.

An Adaptive Damping Network Designed for Strapdown Fiber Optic Gyrocompass System for Ships.

The strapdown fiber optic gyrocompass (strapdown FOGC) system for ships primarily works on external horizontal damping and undamping statuses. When there are large sea condition changes, the system will switch frequently between the external horizontal damping status and the undamping status. This means that the system is always in an adjustment status and influences the dynamic accuracy of the system. Aiming at the limitations of the conventional damping method, a new design idea is proposed, where the adaptive control method is used to design the horizontal damping network of the strapdown FOGC system. According to the size of acceleration, the parameters of the damping network are changed to make the system error caused by the ship's maneuvering to a minimum. Furthermore, the jump in damping coefficient was transformed into gradual change to make a smooth system status switch. The adaptive damping network was applied for strapdown FOGC under the static and dynamic condition, and its performance was compared with the conventional damping, and undamping means. Experimental results showed that the adaptive damping network was effective in improving the dynamic performance of the strapdown FOGC.

A Kalman Filter for SINS Self-Alignment Based on Vector Observation.

In this paper, a self-alignment method for strapdown inertial navigation systems based on the q-method is studied. In addition, an improved method based on integrating gravitational apparent motion to form apparent velocity is designed, which can reduce the random noises of the observation vectors. For further analysis, a novel self-alignment method using a Kalman filter based on adaptive filter technology is proposed, which transforms the self-alignment procedure into an attitude estimation using the observation vectors. In the proposed method, a linear psuedo-measurement equation is adopted by employing the transfer method between the quaternion and the observation vectors. Analysis and simulation indicate that the accuracy of the self-alignment is improved. Meanwhile, to improve the convergence rate of the proposed method, a new method based on parameter recognition and a reconstruction algorithm for apparent gravitation is devised, which can reduce the influence of the random noises of the observation vectors. Simulations and turntable tests are carried out, and the results indicate that the proposed method can acquire sound alignment results with lower standard variances, and can obtain higher alignment accuracy and a faster convergence rate.

AUV Positioning Method Based on Tightly Coupled SINS/LBL for Underwater Acoustic Multipath Propagation.

This paper researches an AUV (Autonomous Underwater Vehicle) positioning method based on SINS (Strapdown Inertial Navigation System)/LBL (Long Base Line) tightly coupled algorithm. This algorithm mainly includes SINS-assisted searching method of optimum slant-range of underwater acoustic propagation multipath, SINS/LBL tightly coupled model and multi-sensor information fusion algorithm. Fuzzy correlation peak problem of underwater LBL acoustic propagation multipath could be solved based on SINS positional information, thus improving LBL positional accuracy. Moreover, introduction of SINS-centered LBL locating information could compensate accumulative AUV position error effectively and regularly. Compared to loosely coupled algorithm, this tightly coupled algorithm can still provide accurate location information when there are fewer than four available hydrophones (or within the signal receiving range). Therefore, effective positional calibration area of tightly coupled system based on LBL array is wider and has higher reliability and fault tolerance than loosely coupled. It is more applicable to AUV positioning based on SINS/LBL.