Tariq P. Sattar

Review: Fight sample degeneracy and impoverishment in particle filters: A review of intelligent approaches

During the last two decades there has been a growing interest in Particle Filtering (PF). However, PF suffers from two long-standing problems that are referred to as sample degeneracy and impoverishment. We are investigating methods that are particularly efficient at Particle Distribution Optimization (PDO) to fight sample degeneracy and impoverishment, with an emphasis on intelligence choices. These methods benefit from such methods as Markov Chain Monte Carlo methods, Mean-shift algorithms, artificial intelligence algorithms (e.g., Particle Swarm Optimization, Genetic Algorithm and Ant Colony Optimization), machine learning approaches (e.g., clustering, splitting and merging) and their hybrids, forming a coherent standpoint to enhance the particle filter. The working mechanism, interrelationship, pros and cons of these approaches are provided. In addition, approaches that are effective for dealing with high-dimensionality are reviewed. While improving the filter performance in terms of accuracy, robustness and convergence, it is noted that advanced techniques employed in PF often causes additional computational requirement that will in turn sacrifice improvement obtained in real life filtering. This fact, hidden in pure simulations, deserves the attention of the users and designers of new filters.

Deterministic resampling: Unbiased sampling to avoid sample impoverishment in particle filters

A novel resampling algorithm (called Deterministic Resampling) is proposed, which avoids uncensored discarding of low weighted particles thereby avoiding sample impoverishment. The diversity of particles is maintained by deterministically sampling support particles to improve the residual resampling. A proof is given that our approach can be strictly unbiased and maintains the original state density distribution. Additionally, it is practically simple to implement in low dimensional state space applications. The core idea behind our approach is that it is important to (re)sample based on both the weight of particles and their state values, especially when the sample size is small. Our approach, verified by simulations, indicates that estimation accuracy is better than traditional methods with an affordable computation burden.

Development of a climbing robot for inspection of long weld lines

Purpose – The project Climbing Robot Cell for Fast and Flexible Manufacture of Large Scale Structures seeks to modernise and take into the future the technology of the manufacture of large fixed welded structures. It creates a transportable manufacturing cell consisting of a team of cooperating climbing robot work tools whose activities are coordinated and integrated through a central intelligence. The purpose of this paper is to present a wheeled robot, called NDT robot, designed for the real time inspection of long weld lines simultaneously with the welding process.Design/methodology/approach – Neodymium permanent magnets are used for adhesion, which are capable of producing a maximum adhesion pressure of 4 × 104 Nm−2 at a 20 mm air gap. The strong neodymium magnets give the robot a high payload carrying capability. The arrangement of the magnet array increases its performance at large air gaps so that the robot has excellent capability to overcome obstacles, such as weld caps. The design of the wheeled...

Design of a climbing robot for inspecting aircraft wings and fuselage

Purpose – The paper aims to develop a robot that climbs on non‐ferrous surfaces, e.g. aircraft wings and fuselages, carrying a heavy payload up to 18 kg including scanning arm and various equipments, for non‐destructive testing (NDT).Design/methodology/approach – This robot in the study uses vacuum suction cups for adhesion, and two pairs of pneumatic cylinders to drive itself, moving in two directions in stepping gait. A rotation mechanism in the centre is used to correct the off‐course deviations by ± 5 degrees. Multiple universal joints are used to make every single suction cup, every robot foot and the whole structure flexible to negotiate with varying surface curvatures presented in different parts of aircraft. This flexible structure is also rigid once the robot is stuck on the surface to enable the NDT inspection being carried out reliably.Findings – The paper finds that the walking speed is limited by the cylinder stroke, time for generating vacuum and changing legs. Although most NDT inspection i...

Adapting sample size in particle filters through KLD-resampling

An adaptive resampling method is provided. It determines the number of particles to resample so that the Kullback-Leibler distance (KLD) between the distribution of particles before resampling and after resampling does not exceed a pre-specified error bound. The basis of the method is the same as Foxs KLD-sampling but implemented differently. The KLD-sampling assumes that samples are coming from the true posterior distribution and ignores any mismatch between the true and the proposal distribution. In contrast, the KLD measure is incorporated into the resampling in which the distribution of interest is just the posterior distribution. That is to say, for sample size adjustment, it is more theoretically rigorous and practically flexible to measure the fit of the distribution represented by weighted particles based on KLD during resampling than in sampling. Simulations of target tracking demonstrate the efficiency of the method.

High-speed Sigma-gating SMC-PHD filter

To solve the general multi-target tracking (MTT) problem, an improved Sequential Monte Carlo (SMC) implementation of the probability hypothesis density (PHD) filter called as Sigma-gating SMC-PHD filter, is proposed that updates particles only using the local nearby measurements inside a specified sigma-gate. The sigma-gate is based on the given measurement noise, e.g. 3@s, where @s is the standard deviation of the measurement noise. Correspondingly, a compensation strategy based on the cumulative distribution function of the measurement model is suggested. Eliminating the contribution of measurements lying outside the gate around the particle will highly reduce unnecessary computation and thus improve the overall processing speed. More importantly, this could shield the estimate from interference from the clutter outside the gate giving more robust and accurate estimation. Especially when the clutter density is high, our approach can yield a win-win that is much faster processing efficiency and better estimation accuracy (as compared with the standard PHD filter). This is demonstrated by simulations of the SMC-PHD filters using measurements of range and bearing, respectively.

Robotic system for inspection of test objects with unknown geometry using NDT methods

The aim of this work is to develop a portable Non-Destructive Testing (NDT) robotic system that can be carried by CLA WAR to evaluate defects in geometrically complex surfaces and industrial products. In this paper, the difference in quality of defect dataduring manual inspection and automated inspection will be compared. Tests have been performed amongst others on complex shaped turbine blades using eddy currents. The challenge is to be able to follow the surface by keeping the NDT probe normal to the surface while maintaining a constant contact force with it. The approach to maintaining a constant contact force and angle to the object was to use a secure contact using permanent magnets and a force adapting control of the manipulator which resulted in improvements inthe quality of inspection compared to manual inspection. Furthermore, a novel application of the position-force-moment (PFM) control was developed. Here, the robotic arm scans unknown contoured surfaces by keeping the sensor probe normal to the test surface, maintaining at the same time constant contact force, thereby ensuring good data acquisition. Copyright

Comparison between MUSIC and ESPRIT direction of arrival estimation algorithms for wireless communication systems

This paper presents simulation of Angle of Arrival (AOA) estimation using the Multiple Signal Classification (MUSIC) and Estimation of Signal Parameters via Rotational Invariance Technique (ESPRIT) algorithms. We study the localization techniques using the wireless communication systems and there are several algorithms that have the ability in calculate the Direction of Arrival (DOA) of the incident signals. We investigate and compare MUSIC and ESPRIT algorithms. The simulations indicate that the MUSIC algorithm is more accurate and stable compared to the ESPRIT algorithm.

Amphibious NDT Robots

Oil, petrochemical, and food processing industries worldwide store their raw materials and product in tens of thousands of storage tanks. The tanks are mostly constructed using welded steel plates and therefore subject to corrosion and weld cracking. Testing the structural integrity of these storage tanks with non-destructive testing (NDT) techniques is an expensive and time consuming activity. The walls of a large tank can usually be tested manually (for corrosion thinning and weld defects using ultrasonic techniques) from outside the tank. Access to most areas of a wall is obtained by constructing scaffolding or abseiling down from the top. However, erecting scaffolding is expensive and the inspection is tedious and slow. These costs can be reduced and the inspection speeded up by using climbing robots that deploy ultrasonic probes with scanning arms. However, there are some areas of the wall that cannot be accessed from the outside, for example at the base where the walls are protected by striker plates to run off rain water, or behind wind girders around the tank that are used to strengthen the tank, or obviously in tanks that are partially or fully buried in the ground. Similarly, storage tanks on ships cannot be inspected from the outside because they are surrounded by ballast tanks. To inspect the walls in these cases, the inspection has to be performed from inside the tank. Additionally, a primary source of trouble in tanks is the floor which can corrode quickly in localised areas due to ground chemistry and suffer damage to welds through soil movement. The floors can be inspected reliably only by entering the tank. Entry into tanks can be performed only when the tank has been emptied and thoroughly cleaned when it contains hazardous materials. The operation is hugely expensive requiring transportation of product to other tanks or locations, outages that last many months with subsequent loss of revenue, and cleaning costs incurred by repeated cleaning of the tank till all product is removed. Very large cost savings can be made by performing internal and in-service inspection of the walls and floors of storage tanks by using robots that can be inserted into a tank through manholes. The robots will have to operate in the product contained in the tank (e.g. crude oil, refined petroleum products, chemicals, liquors, etc.) and the robot should be capable of gaining access to target areas that are to be inspected (tank wall areas, floor areas, and welds on the walls or floor). O Access Database www.i-techonline.co m

Novel Solutions to Design Problems of Industrial Climbing Robots

This paper aims to identify some design problems that affect system performance of climbing robots and to present novel solutions to the problems. The first problem is the interference of the umbilical cord for the robot with its mobility and dynamics; the second problem is the slow traveling speed of climbing robots, affecting the overall working efficiency of the system; the third one is unforeseen variations of surface curvature of objects on which a vacuum suction type of robot climbs, which reduces suction pressures and adhesion forces pulling the robot onto the surface; and the fourth is the intrinsic safety requirements of electrical components in flammable environments. The technical solutions have been developed and tested in laboratory and industrial field trials and have been proven to be well founded and unique. It has been verified that new techniques from other engineering areas, e.g. materials and information engineering, can be used to formulate novel solutions to overcome the obstacles. The technical approaches used in these solutions will strengthen the state of the art and contribute to the advance of climbing robot applicable technology.