A hierarchical multiple-model approach for detection and isolation of robotic actuator faults

Modern robotic systems perform elaborate tasks in complicated environments and have close interactions with humans. Therefore fault detection and isolation (FDI) schemes must be carefully designed and implemented on robotic systems in order to guarantee safe and reliable operations. In this paper, we propose a hierarchical multiple-model FDI (HMM-FDI) scheme to detect and isolate actuator faults of robot manipulators. The proposed algorithm performs FDI in stages and refines the associated model set at each stage. Consequently only a small number of models are required to detect and isolate various types of unexpected actuator faults, including abrupt faults, incipient faults, and simultaneous faults. In addition, the computational load is alleviated due to the reduced-sized model set. The relation between the fault detection stage of the HMM-FDI scheme and the likelihood ratio test is explicitly revealed and theoretical upper bounds of the false alarm and missed detection probabilities are evaluated. Then we conduct experiments to demonstrate the ability of the HMM-FDI scheme in successful and immediate detection and isolation of actuator faults.     

A pulsed plasma thruster fault detection and isolation strategy for formation flying of satellites

The main objective of this paper is to develop a dynamic neural network-based fault detection and isolation (FDI) scheme for pulsed plasma thrusters (PPTs) that are employed in the attitude control subsystem (ACS) of satellites tasked to perform formation flying (FF) missions. A hierarchical methodology is proposed that consists of three fault detection and isolation (FDI) approaches, namely (i) a “low-level” FDI scheme, (ii) a “high-level” FDI scheme, and (iii) an “integrated” FDI scheme. Based on the data from the electrical circuit of the PPTs, the proposed “low-level” FDI scheme can detect and isolate faults in the PPT actuators with a good level of accuracy, however the precision level is poor and below expectations with the misclassification rates as expressed by False Healthy and False Faulty parameters being too high. The proposed “high-level” FDI scheme utilizes data from the relative attitudes of the FF mission. This scheme has good detection capabilities, however its isolation capabilities are not adequate. Finally, the proposed “integrated” FDI scheme takes advantage of the strengths of each of the above two schemes while reducing their individual weaknesses. The results demonstrate a high level of accuracy (99.79%) and precision (99.94%) with a misclassification rates that are quite negligible (less than 1%). Furthermore, the proposed “integrated” FDI scheme provides additional and interesting information related to the effects of faults in the thrust production levels that would not have been available from simply using the low or the high level schemes alone.     

Cascade filter structure for sensor/actuator fault detection and isolation of satellite attitude control system

This paper presents a new scheme for fault detection and isolation in a satellite system. The purpose of this paper is to develop detection, isolation and identification algorithms based on a cascade filter for both total and partial faults in a satellite attitude control system (ACS). The cascade filter consists of a decentralized Kalman filter (DKF) and a bank of interacting multiple model (IMM) filters. The cascade filter is utilized for detection and diagnosis of anticipated sensor and actuator faults in a satellite ACS. Other fault detection and isolation (FDI) schemes are compared with the proposed FDI scheme. The FDI procedure using a cascade filter was developed in three stages. In the first stage, two local filters and a master filter detect sensor faults. In the second stage, the FDI scheme checks sensor residuals to isolate sensor faults, and 11 Extended Kalman filters with actuator fault models detect wherever actuator faults occur. In the third stage of the FDI scheme, four filters identify the fault type, which is either a total or partial fault. An important feature of the proposed FDI scheme is that it can decrease fault isolation time and accomplish not only fault detection and isolation but also fault type identification using a scalar penalty in the conditional density function.     

A comparison of model-based path control algorithms for direct-drive SCARA robots

Various advanced control strategies are applied to a direct-drive SCARA robot and studied in computer simulations. Besides computed torque control and direct adaptive control, heuristic optimal control, a new path control scheme for robotic manipulators, is included in the comparison study. PD control, the traditional robot control method, is used for generating a comparing baseline. While all schemes are applied for the same tracking task, the effect of modelling errors and measurement noise is considered in robot performance evaluation. Simulation results show that (1) without model errors, all advanced control schemes can achieve higher tracking accuracy than PD control; (2) with a random measurement error of 1%, computed torque and direct adaptive control methods are inferior to PD control; (3) heuristic control proves to be the most robust control scheme in case of mixed model and measurement errors.     

Motion coordination and reactive control of autonomous multi-manipulator systems

The emerging field of service robots demands new systems with increased flexibility. The flexibility of a robot system can be increased in many different ways. Mobile manipulation—the coordinated use of manipulation capabilities and mobility—is an approach to increase robots flexibility with regard to their motion capabilities. Most mobile manipulators that are currently under development use a single arm on a mobile platform. The use of a two-arm manipulator system allows increased manipulation capabilities, especially when large, heavy, or non-rigid objects must be manipulated. This article is concerned with motion control for mobile two-arm systems. These systems require new schemes for motion coordination and control. A coordination scheme called transparent coordination is presented that allows for an arbitrary number of manipulators on a mobile platform. Furthermore, a reactive control scheme is proposed to enable the platform to support sensor-guided manipulator motion. Finally, this article introduces a collision avoidance scheme for mobile two-arm robots. This scheme surveys the vehicle motion to avoid platform collisions and arm collisions caused by self-motion of the robot. © 1996 John Wiley & Sons, Inc.     

Redundancy modelling and resolution for robotic mobile manipulators: a general approach

In this work the topic of kinematic redundancy modelling and resolution for robotic mobile manipulators is considered. A set of redundancy parameters is introduced to define a general inverse kinematic procedure for mobile manipulators. Then, redundancy is treated as a non-linear optimization problem with the purpose of finding robot configurations that maximize the designed metric measures. Some strategies to design the optimization objective function are introduced in order to achieve desirable redundant behaviours, such as obstacles avoidance, mobile base motions reductions and dexterity optimization. Moreover, the robot controller has been developed following an object-oriented software architecture principle that allows to keep it general and robot independent. As a prove of reliability and generality of our approach, the same controller has been used to control several different mobile manipulators in a simulation environment, as well as a real KUKA youBot robot.     

A fully adaptive decentralized control of robot manipulators

In this paper, we develop a fully adaptive decentralized controller of robot manipulators for trajectory tracking. With high-order and adaptive variable-structure compensations, the proposed scheme makes both position and velocity tracking errors of robot manipulators globally converge to zero asymptotically while allowing all signals in closed-loop systems to be bounded, even without any prior knowledge of robot manipulators. Thus this control scheme is claimed to be fully adaptive. Even when the proposed scheme is modified to avoid the possible chattering in actual implementations, the overall performance will remain appealing. Finally, numerical results are provided to verify the effectiveness of the proposed schemes at the end.     

Robust control of robot manipulators: A survey

This paper presents an overview of robust control schemes for robot manipulators. The survey summarizes the vast literature on the subject. The different modelling assumptions used in current control algorithms are thoroughly reviewed. The survey includes models of actuator dynamics and joint flexibility. The different control schemes are organized in the following six categories: linear schemes, passivity-based schemes, Lyapunov-based schemes, sliding mode control schemes, non-linear H omega schemes and robust adaptive control schemes. Connections and comparisons are made between the various algorithms.     

Finding fault - fault diagnosis on the wheels of a mobile robot using local model neural networks

As mobile robots are mostly designed to act autonomously, procedures that detect and isolate faults on the various parts of a robot are essential. The most powerful approaches in fault detection and isolation (FDI) are those using a process model, where quantitative and qualitative knowledge-based models, databased models, or combinations thereof are applied. This article suggests a model-free approach to the solution of the fault detection problem. One way to deal with the absence of a mathematical model is to build a model from input-output data. In this article, local model networks (LMNs) are used for plant modeling. The key to fault detection and diagnosis is the creation of residual signals. Although the way these signals are formed varies, in all cases the residuals change their value accordingly with the presence of faults. To avoid false alarms, the residuals must be affected by factors unrelated to faults (like modeling errors) as little as possible. Change-detection algorithms are therefore used for reliable residual generation. These algorithms are designed to detect changes in signals that include noise or other types of disorders. The combination of local model networks for modeling and change-detection algorithms for residual creation provides an efficient method for fault detection and diagnosis. The method is applied on the wheels subsystem of a mobile robot.     

An overview of collaborative robotic manipulation in multi-robot systems

Robotic manipulation aims at combining the versatility and flexibility of mobile robot platforms with manipulation capabilities of robot manipulators. This survey paper comprehensively reviews the state-of-the-art development of collaborative robotic manipulation from the perspective of modelling, control and optimization. Then, the recent results in this field can be categorized into coordination of multiple fixed manipulators, mobile robots and mobile manipulators, respectively. A classification and comparison of various issues and promising approaches is given. Finally, a short discussion section is given to summarize existing research and to point out several future research directions.     

机械手的线性扰动和模型跟随自适应控制方案的仿真研究

本文提出了机械手在关节坐标系中的引入积分作用的线性扰动自适应控制和引入重力补偿的模型跟随自适应控制方案.在 PC/XT 机上的仿真结果表明,上述两个控制方案,即使在负载未知时也能使机械手跟随期望轨迹.     

Neural-network-based robust fault diagnosis in robotic systems

Fault diagnosis plays an important role in the operation of modern robotic systems. A number of researchers have proposed fault diagnosis architectures for robotic manipulators using the model-based analytical redundancy approach. One of the key issues in the design of such fault diagnosis schemes is the effect of modeling uncertainties on their performance. This paper investigates the problem of fault diagnosis in rigid-link robotic manipulators with modeling uncertainties. A learning architecture with sigmoidal neural networks is used to monitor the robotic system for any off-nominal behavior due to faults. The robustness and stability properties of the fault diagnosis scheme are rigorously established. Simulation examples are presented to illustrate the ability of the neural-network-based robust fault diagnosis scheme to detect and accommodate faults in a two-link robotic manipulator.     

Fault diagnosis of a class of nonlinear uncertain systems with Lipschitz nonlinearities using adaptive estimation

This paper presents a fault detection and isolation (FDI) scheme for a class of Lipschitz nonlinear systems with nonlinear and unstructured modeling uncertainty. This significantly extends previous results by considering a more general class of system nonlinearities which are modeled as functions of the system input and partially measurable state variables. A new FDI method is developed using adaptive estimation techniques. The FDI architecture consists of a fault detection estimator and a bank of fault isolation estimators. The fault detectability and isolability conditions, characterizing the class of faults that are detectable and isolable by the proposed scheme, are rigorously established. The fault isolability condition is derived via the so-called fault mismatch functions, which are defined to characterize the mutual difference between pairs of possible faults. A simulation example of a single-link flexible joint robot is used to illustrate the effectiveness of the proposed scheme.     

Fault detection and identification based on combining logic and model in a wall-climbing robot

A combined logic- and model-based approach to fault detection and identification （FDI） in a suction foot control system of a wall-climbing robot is presented in this paper. For the control system, some fault models are derived by kinematics analysis. Moreover, the logic relations of the system states are known in advance. First, a fault tree is used to analyze the system by evaluating the basic events （elementary causes）, which can lead to a root event （a particular fault）. Then, a multiple-model adaptive estimation algorithm is used to detect and identify the model-known faults. Finally, based on the system states of the robot and the results of the estimation, the model-unknown faults are also identified using logical reasoning. Experiments show that the proposed approach based on the combination of logical reasoning and model estimating is efficient in the FDI of the robot.     

重复运动速度层和加速度层方案的等效性

探讨冗余机械臂之重复运动规划(Repetitive motion planning, RMP), 对应速度层重复运动规划方案, 提出了加速度层重复运动规划方案, 并进一步发现和探讨了速度层与加速度层RMP方案的等效性关系. 基于神经动力学方法的理论分析和基于PUMA560机械臂的计算机仿真结果均证实了本文所提出的加速度层RMP方案在冗余度解析问题上的有效性, 以及速度层和加速度层RMP方案的等效性.     

The merits of exergy-based fault detection in petrochemical processes

The complexity and multi-domain nature of petrochemical (PC) plants make the application of conventional model-based fault detection and isolation (FDI) techniques a challenging endeavour. Although hybrid FDI schemes aim to address this shortfall, many are simply a combination of data-driven techniques that exclude physical system information. In this work, a hybrid approach to FDI of a PC process is proposed that is based on an exergy-data abstraction. Data from an actual system is abstracted to system exergy, based on physical knowledge of the system and then used as a diagnostic metric for the FDI scheme. In this paper, it is shown why energy-based approaches are lacking when considering petrochemical processes. After presenting a novel method for the real-time, automatic calculation of chemical exergy in Aspen HySys^® the applicability of exergy-based fault detection is investigated. Application of the exergy-based fault detection scheme shows a marked improvement over the energy-based approach with perfect detectability and isolability of the considered process faults. The exergy-based fault detection technique shows merit in comparison to the energy-based detection scheme. Additionally, and more importantly, exergy-based characterisation allows the use of more sophisticated model-based fault detection schemes to petrochemical processes.     

Fault diagnosis of an industrial gas turbine prototype using a system identification approach

In this work, a model-based procedure exploiting analytical redundancy for the detection and isolation of faults on a gas turbine simulated process is presented. The main point of the paper consists of exploiting an identification scheme in connection with dynamic observer or filter design procedures for diagnostic purposes. Thus, black-box modelling and output estimation approaches to fault diagnosis are in particular advantageous in terms of solution complexity and performance achieved. Moreover, the suggested scheme is especially useful when robust solutions are considered for minimising the effects of modelling errors and noise, while maximising fault sensitivity. In order to experimentally verify the robustness of the solution obtained, the proposed FDI strategy has been applied to the simulation data of a single-shaft industrial gas turbine plant in the presence of measurement and modelling errors. Hence, extensive simulations of the test-bed process and Monte Carlo analysis are the tools for assessing experimentally the capabilities of the developed FDI scheme, when compared also with different data-driven diagnosis methods.     

On the adaptive control of robot manipulators with velocity observers

To achieve accurate tracking control of robot manipulators many schemes have been proposed. A common approach is based on adaptive control techniques, which guarantee trajectory tracking under the assumption that the robot model structure is perfectly known and linear in the unknown parameters, while joint velocities are available. Despite tracking errors tend to zero, parameter errors do not unless some persistent excitation condition is fulfilled. There are few works dealing with velocity observation in conjunction with adaptive laws. In this note, an adaptive control/observer scheme is proposed for tracking position of robot manipulators. It is shown that tracking and observation errors are ultimately bounded, with the characteristic that when a persistent excitation condition is matched then they, as well as the parameter errors, tend to zero. Simulation results are in good agreement with the developed theory.     

An optimized Takagi-Sugeno type neuro-fuzzy system for modeling robot manipulators

The present paper describes the development of a Takagi-Sugeno (TS)-type Neuro-fuzzy system (NFS) for dynamic modeling of robot manipulators. The NFS has been trained by a relatively new combinatorial metaheuristic optimization method, called particle swarm optimization (PSO). The development of such an intelligent, robust, dynamic models for robot manipulators can immensely help in deriving proper position/velocity control strategies in offline situations with these accurately developed models. The proposed PSO-based NFS has been successfully applied to two-link and three-link model robot manipulators.     

Highly efficient transputer arrays for the computation of robot dynamics

The modelling of the highly coupled and non-linear dynamic models of robot manipulators is computationally expensive in terms of computer hardware. Hence this problem has always presented a major obstacle in on-line dynamic control applications. However, the recent advent of the VLSI single on-chip computer (Transputer) makes it possible to implement new algorithms that compute these dynamic models within real-time constraints. This paper describes a solution of this problem by employing a parallel-processing approach. The dynamic model of a robot manipulator is divided into different tasks. Further, each task is divided into several subtasks. A heuristic scheduling algorithm is used to produce different near-optimum task allocations. Real-time implementations of the proposed tasks allocations are given to demonstrate the efficiency and superiority of the parallel-processing approach.