Informatics Publishing Limited

A. Sampath Dakshina Murthy ¹, S. Koteswara Rao ², K. Srinavasa Naik ¹, Rudra Pratap Das ¹, Kausar Jahan ¹, K. Lova Raju ¹

Affiliations
1 Department of ECE, Vignan’s Institute of Information Technology, Visakhapatnam - 530046, Andhra Pradesh, IN
2 Department of ECE, K L University, Vaddeswaram, Guntur District - 522502, Andhra Pradesh, IN

Abstract

For on sea water measurements of range on bearing in naval operations various types of Kalman filters including extended Kalman filter have been used. In this paper ,input estimation techniques applied normally to in air applications are applied to on sea measurements using chi – square distribution random sequence residual techniques. In which the algorithm estimates target motion parameters covering maneuvering. After detecting the target maneuvering, the algorithm corrects the velocity and position components based on acceleration noise input estimation method. Monte – Carlo simulation has proved the efficiency of the method. Practical applications of these methods are quite immense. Further test can be carried out on such cases.

Keywords

Chi-Square Distribution, Extended Kalman Filter, Linear Maneuvering, Monte-Carlo Simulator, Radar Measurements

Full Text

   

K. Lova Raju ¹, S. Koteswara Rao ², Rudra Pratap Das ¹, M. Nalini Santhosh ³, A. Sampath Dakshina Murthy ¹

Affiliations
1 Department of ECE, Vignan’s Institute of Information Technology, Visakhapatnam - 530046,Andhra Pradesh, IN
2 Department of ECE, K L University, Vaddeswaram, Guntur District - 522502, Andhra Pradesh, IN
3 Department of ECE, Vignan’s Institute of Information Technology, Visakhapatnam - 530046, Andhra Pradesh, IN

Abstract

In surveillance sonar passive tracking for bearing measurements has been done so far by using kalman filter and its modifications. In this paper, target is assumed to be moving mostly at constant velocity in underwater scenario. Unscented Kalman Filter has been utilised to track underwater submarine target/moving ship. Monte Carlo Simulation in PC environment is used to prove the effectiveness of UKF. Cases taken up are Ownship/Target S or L manoeuvring. Methods applied include for target tracking, mathematical modelling, and bearings only techniques. Simulation results have been obtained and comparison has been made based on typical scenarios. Different configurations of underwater scenarios can be taken up for further research.

Keywords

Bearings, Line of Sight, Monte Carlo Simulation, Ownship, S-Manoeuvring, Unscented Kalman filter

Full Text

   

Jawahar Annabattula ¹, S. Koteswara Rao ¹, A. Sampath Dakshina Murthy ², K. S. Srikanth ¹, Rudra Pratap Das ²

Affiliations
1 School of Electrical Sciences, KLEF, KL University, Guntur - 522502, Andhra Pradesh, IN
2 Department of ECE, Vignan’s Institute of Information Technology, Visakhapatnam - 530046, Andhra Pradesh, IN

Abstract

Background/Objectives: In underwater, a single submarine processes noisy sonar bearings from a radiating target (ship/submarine/torpedo) in passive listening mode. Methods/Statistical Analysis: The target is assumed to be at constant velocity or maneuvering in its course/speed/course and speed. An adaptive Kalman filter estimates target motion parameters. Findings: Data fusion and data association in multi target and multi sensor scenario are carried out, using Chi-squared innovation process. Applications/Improvements: Finally improved torpedo intercept guidance algorithm with homing capabilities is developed for attacking the target.

Keywords

Data Fusion, Doppler, Fire Control System, Towed Array

Full Text

   

Jawahar Annabattula ¹, S. Koteswara Rao ¹, A. Sampath Dakshina Murthy ², K. S. Srikanth ¹, Rudra Pratap Das ²

Affiliations
1 School of Electrical Sciences, KLEF, KL University, Guntur - 522502, Andhra Pradesh, IN
2 Department of ECE, Vignan’s Institute of Information Technology, Visakhapatnam - 530046, Andhra Pradesh, IN

Abstract

Background/Objectives: The modern submarines use mutlitple sensors for tracking multiple targets in sea environment. In general, multiple sensor data can be handled in two ways: measurement fusion and state vector fusion. Methods/ Statistical Analysis: Measurement fusion is not practical for implementation due to various reasons and hence state vector fusion is proposed. Target trackingusing Modified Gain Bearings only Extended Kalman Filter in generic two-dimensional platform is carried out in each channel. Findings: In this approach, a state vector and its corresponding covariance matrix are extracted from the sensor measurements by an estimator equipped on each sensor. The output of each channel is transported via a data link in to order to be reach the fusion center. Applications/Improvements: A composite target state vector is obtained by performing track-to-track correlation and fusion at the fusion center.Monte-Carlo simulation is carried outand the results are presented for a typical scenario.

Keywords

Estimation, Fusion, Kalman Filter, Simulation, Tracking

Full Text

   

Jawahar Annabattula ¹, S. Koteswara Rao ¹, A. Sampath Dakshina Murthy ², K. S. Srikanth ¹, Rudra Pratap Das ²

Affiliations
1 School of Electrical Sciences, KLEF, KL University, Guntur - 522502, Andhra Pradesh, IN
2 Department of ECE, Vignan’s Institute of Information Technology, Visakhapatnam - 530046, Andhra Pradesh, IN

Abstract

In underwater scenario, observer manoeuvre is required to find out target motion parameters in bearings only passive target tracking. Sometimes, due to tactical constraints, observer is not able to carry out manoeuvre. In this paper, it is shown that target motion parameters can be obtained in such situation, if the knowledge of any one of the target motion parameters is available. There are other practical problems like spurious bearings are generated by sonar, etc. In addition, auto tracking fails often and some bearings will be missed. Pseudo Linear Kalman Filter is made flexible to address these practical problems.

Keywords

Constrained Environment, Direction of Arrival Estimation, Kalman Filtering, Monte Carlo Simulation, Passive Target Tracking, Target Detection

Full Text

   

T. Pavani ¹, Rudra Pratap Das ¹, A. Naga Jyothi ¹, A. Sampath Dakshina Murthy ¹

Affiliations
1 Department of ECE, Vignan’s Institute of Information and Technology, Visakhapatnam – 530049, Andhra Pradesh, IN

Abstract

Linear and circular antenna arrays need to be properly designed for optimization of electromagnetic problem. The design parameters are decided by the specific beam width using firefly algorithm, desired radiation patterns are to be studied for the reduction of beam width and side lobe level simultaneously based on multi objective approach.PSO is used for comparison based on simulation examples. Linear antennas are designed to obtain the source locations in position only synthesis based on FA. Patterns are synthesized using circular arrays to achieve minimum side lobe load and high directivity with constraint of beam width. Performance evaluation of bat, cuckoo search algorithms is level as compared to FA. This method is extended to concentric circular arrays with dipoles as radiating elements, FA is found to be most suitable for development of linear and circular antenna arrays .Procedure of thinning helps in reducing the cost of array designs. This method will find extensive usage in design of array pattern. PSO method of comparison can be supplemented with GA for better results.

Keywords

Particle Swarm Optimization, Firefly Algorithm, Cuckoo Search Algorithms, Circular Antenna, Bat Algorithm, Sidelobe Level, Beamwidth.

Full Text

   

A. Sampath Dakshina Murthy ¹, S. Koteswara Rao ², G. Thiagarajan ¹, V. Suresh ³

Affiliations
1 Department of ECE, Vignan’s Institute of Information Technology, Beside VSEZ, Duvvada, Gajuwaka, Visakhapatnam - 530046, Andhra Pradesh, IN
2 Department of ECE, K L University, Green Fields, Vaddeswaram, Guntur District - 522502, Andhra Pradesh, IN
3 Department of CSE, Vignan’s Institute of Information Technology, Beside VSEZ, Duvvada, Gajuwaka, Visakhapatnam - 530046, Andhra Pradesh, IN

Abstract

For processing signals and in control application filters are essential, linear optimums discrete time filters such as wiener filter and Kalman filter are on orthogonal principle. For non stationary cases of having a presence of noise, adaptive wiener filter has to be applied using Monte Carlo Simulation 250 samples were used for 50 runs. Coefficients of linear filter are used to estimate the additive white noise. Error is calculated and RMS value of each error is added to the sample for desired signal. FIR wiener filters of order 6, 12, 24 were chosen for adaptive operators .Simulation results were quite encourage in the sense that noise was suppressed to maximum extends. Adaptive methods noisy higher number of samples and more than 100 runs are linear to yield better results.

Keywords

Noise Cancellation, Monte Carlo Simulation, Wiener Filter, Optimal Filter, Wiener-Hopf Equations, Wide-Sense Stationary Random Processes, Discrete Wiener Filter, Discrete Kalman Filter.

Full Text

   

A. Sampath Dakshina Murthy ¹, S. Koteswara Rao ², A. Naga Jyothi ¹, Rudra Pratap Das ¹

Affiliations
1 Department of ECE, Vignan’s Institute of Information Technology, Visakhapatnam – 531162, Andhra Pradesh, IN
2 Department of ECE, K. L. University, Vaddeswaram, Guntur District – 522502, Andhra Pradesh, IN

Abstract

In one dimensional vertical motion in the case of tracking a target, many types of filters included Kalman varieties have been used so far. This paper is presenting a new approach of inverse ballistic coefficient included in the target state vector for reducing the convergence time. For analysis, estimation of object’s altitude and velocity is done by simulation based on noisy error environment measurements. The important of beta (ballistic coefficient) for non linear studies on extended Kalman filter has been established with proper evidence. This work can be further extended to consider the effect of hang off errors and air drag.

Keywords

Air Drag, Altitude, Beta Coefficient, Extended Kalman Filter, Hang Off Error, Inverse Beta, Velocity

Full Text