In this paper, the problem of combined vision/force servo control for robot manipulator is addressed. Three different robot vision control strategies: position-based, image-based, and hybrid control are combined with an impedance-based force controller and a comparison of these three combined vision/force control methods is investigated for the first time, in the context of generic robot kinematic-based sensory-task-space control structure. Furthermore, the issue of contact surface parameters estimation is also investigated. Simulation results have demonstrated that all the above vision/force control strategies are comparable in terms of both the dynamic response and accuracy of positioning and force control.

IC chip has gradually become smaller and smaller, and thus it requires high packaging density. In chip packaging, accurate alignment of electronic components with respect to PCB is crucial for high quality packaging, especially in flipchip assembly. In this paper, vision system is used to provide relative pose information between flip-chip and substrate. Based on this information, these two parts are aligned accurately using visual servoing. In order to achieve high accuracy alignment, a dual imaging system (DIS) is introduced in this work, which is composed of zoom lenses, beam-splitter, mirror, CCD, and LED illumination. It can simultaneously observe the solder bumps on flip-chip and the pattern of pads on substrate using one camera. Once the image frame containing flip-chip and substrate is obtained, their features are extracted from the preprocessed image. Extraction of the features enables us to obtain the position and orientation errors between the chip and the substrate. On the base of the measured errors, visual servoing method can determine the instantaneous velocity input of flip-chip at each servoing time and control the relative position and orientation precisely in an on-line manner. We carry out a series of experiments for various magnifications in order to evaluate the performance of the dual imaging system and the visual servoing algorithm as well.

Visual servoing technique used in microassembly limits its application, due to the two inherent problems, small depth of focus and small field of view. To avoid this criticism, in this paper, we present a method of microassembly using active zooming, which can solve these problems. In order to solve the problem of small field of view, the method using active zooming that can prevent the target out of field of view is proposed. In order to solve the problem of the small depth of focus, we proposed a method to control the zooming lens based on the blur measure of the moving target to get the clear image in the field of view. In this paper, we use affine transformation of feature line to track the object. The proposed microassembly process consists coarse visual servoing and fine visual servoing. A series of experiments are performed on peg and hole assembly to investigate the feasibility of this method.

In this paper, a new open architecture for visual servo control tasks is illustrated. A Puma-560 robotic manipulator is used to prove the concept. This design enables doing hybrid force/visual servo control in an unstructured environment in different modes. Also, it can be controlled through Internet in teleoperation mode using a haptic device. Our proposed structure includes two major parts, hardware and software. In terms of hardware, it consists of a master (host) computer, a slave (target) computer, a Puma 560 manipulator, a CCD camera, a force sensor and a haptic device. There are five DAQ cards, interfacing Puma 560 and a slave computer. An open architecture package is developed using Matlab^(R), Simulink^(R) and XPC target toolbox. This package has the Hardware-In-the-Loop (HIL) property, i.e., enables one to readily implement different configurations of force, visual or hybrid control in real time. The implementation includes the following stages. First of all, retrofitting of puma was carried out. Then a modular joint controller for Puma 560 was realized using Simulink^(R). Force sensor driver and force control implementation were written, using sfunction blocks of Simulink^(R). Visual images were captured through Image Acquisition Toolbox of Matlab^(R), and processed using Image Processing Toolbox. A haptic device interface was also written in Simulink^(R). Thus, this setup could be readily reconfigured and accommodate any other robotic manipulator and/or other sensors without the trouble of the external issues relevant to the control, interface and software, while providing flexibility in components modification.

In this paper a new method is proposed to control a vision-based robot in large navigation spaces. In this case, visual features observed by an on-board camera can change drastically or even disappear completely between the initial image, as seen at the beginning of a task, and the final image, as seen at the desired position of the robot. These features are therefore not suffcient for controlling the entire motion of the robotic system from beginning to end. This problem requires a more complete definition and representation of the navigation space. This can be achieved by a topological representation, where the environment is directly defined in the sensor space by a data-base of images. In our approach, this data-base is acquired during an offline learning step. An image retrieval method then indexes and matches a request image, given by the camera, to the closest view within the data-base. In this way, an image path is extracted from the database to link the initial and desired images providing enough information to control the robot. The central point of this paper is focused on the closed-loop control law that drives the robot to its desired position using this image path. The method proposed does not require either a global reconstruction or a temporal planning step. Furthermore, the robot is not obliged to converge directly upon each image waypoint but chooses automatically a better trajectory. The visual servoing control law designed uses specific features which ensure that the robot navigates within the visibility path. Experimental simulations are given to show the effectiveness of this method for controlling the motion of a camera in three-dimensional environments (free-flying camera, or camera moving on a plane).

Design of objects that are used in vision-guided robotic systems crucially affects the overall system performance. In this paper, we target the problem of optimal feature point design for a given camera motion profile in robotic eye-in-hand systems. Having the intrinsic camera calibration parameters, the motion profile, and the image Jacobian matrix, a new directional relative motion resolvability measure is introduced. For each known point on the camera trajectory with known camera-to-object relative pose, the proposed measure is evaluated as a separate objective, resulting in a multi-objective problem. A bounded multi-objective optimization approach is successfully utilized to solve the underconstrained feature design problem. Simulation results show that the motion of the camera is better resolved for the optimally designed object.

We propose self-maintenance robot system as a method which realizes work for a long time without maintenance by the human workers. This system absorbs the change which occurs in robot's hardware by learning, and maintains working ability. We propose the two methods of learning changes in the physical information of the robot as methods which realizes the maintenance-free robot system. One is a method to learn robot's physical information based on the input and output information in the task practice from the no physical information of the robot by using a neural network which has a task common layer and a task independence layer. We use a neural network which has a task common layer and a task independence layer to learning. Other is a method to learn robot's physical information based on the difference in hoping action and actual action. In this report, we verify of these learning system by the computer simulation.

The ongoing Chinese ever-ambitious project of Large Sky Area Multi-Object Fiber Spectroscopic Telescope (LAMOST) has brought about a tremendous challenge for the control engineers. To the bottom line the giant 4-meter class ground telescope is a comprehensive optomechatronic platform to achieve high performance and functionality, such as its capability of observing 4000 stars simultaneously, which will set a world record in contemporary ground survey telescopes. This paper outlines the R&D stages of the control system for the project along with its integrated strategy of optomechatronic components in general and network control framework in particular. The approach is to make a careful investigation with respect to the time crucialness for execution of different tasks so as to utilize different networks. However, the overall network framework is based on a distributed platform, hierarchical structure and open architecture to boost the flexibility. Vigorous study has been invested and a number of cutting edge techniques have been applied to meet the tough network control requirements, such as real-time database, powerful interfaces, sophisticated controllers, remote control, etc.

Inertial stabilization of electro-optical sighting systems and weapon slaving control loops are essential constituents of modern fire control systems for mobile combat vehicles. These systems are used for surveillance, target tracking and engaging the targets under dynamic conditions. Firing accuracy of such systems largely depends on stabilization and weapon slaving accuracies. Accuracy requirements become stringent as the operating range increases. Several other issues such as bore sighting offsets, ballistic offsets and mounting error compensation etc. are also to be considered. Fuzzy knowledge based controller (FKBC) offers an alternative method to the conventional control synthesis methodologies using root locus, Bode plots or pole placement. Fuzzy control loops are particularly useful when the plant consists of substantial non-linearity due to actuator saturation, stiction, Coulomb friction, digitization etc. Since, the control surface obtained through this method is non-linear, generally it provides greater flexibility to designer to achieve better damping, lesser control energy even in presence of various constraints. This work presents the design of weapon slaving loop using a fuzzy controller. The weapon is slaved to a gimbaled electro-optical sight, which has a stabilized line of sight along two axes. The system under consideration is designed for naval platforms. A two-input (error and rate of change of error) and single output (incremental control) fuzzy controller has been designed to position the weapon at desired position. Implementation of controller has been done using digitized inputs. Simulations have been carried out to evaluate the performance of the integrated fire control system under the presence of various non-linearities, sensor inaccuracies and other exogenous inputs like host platform generated disturbances and measurement noise. Stringent requirements of disturbance attenuation, tracking and command following have been met.

Design strategy for linear parameter varying systems (LPV) is considered in the two-degree-of-freedom control framework. Firstly, coprime factorization described in state space formulas for LPV systems are newly introduced based on parameter-dependent Lyapunov function. Secondly, based on this coprime factorization, two-degree-of-freedom control framework of LTI system is extended to LPV systems. Good tracking performance and good disturbance rejection are compatibly achieved by a feedforward controller and a feedback controller, respectively. Furthermore, each controller design problem can be formulated in terms of linear matrix inequality related to L2 gain performance. Finally, a simple design example is illustrated.

This paper introduces tracking control system with visual feedback to a moving object by using the measurement device which we developed. In order to recognize the moving object, we use two method, using cross-shape mark and Orientation Code Matching (OCM). And the measurement device is constructed PID control system with Extended Kalman Filter in order to track to object. Through the several experiments, we verify the percormance of recognition and tracking.

In this paper, we investigate effect of changing the connection of feed-forward loop based on error signal. Our motivation of this work is solution to progress of human skill. For the skill model, we study a human simple action such as arm motion. Many models that describe the human arm dynamics have been proposed in recent year. While one type does not need an inverse model of human dynamics, the system based on the model does not include feed-forward loop. On the other hand, another type model has a feed-forward loop and feedback loop systems. This type assumes feed-forward element includes an internal model by repeating action or training and this loop progress our skill. Then we usually have to exercise to get a good performance. This says that we design the internal motion model by training and we move on prediction for motion. Under the assumption, Kawato model is well known. The model proposed that learning of feed-forward element is promoted in brain so that the error of feedback loop decreases. Furthermore, we assume the connections in feedback loop and feed-forward loop are changed. We show numerical simulations and consider that the position error given by our vision changes the skill element and we confirm that the position error is the one of the estimate function for the improvement in our skill.

In this paper, a linear robust control system design based on μ-synthesis is proposed for impedance shaped 2-DOF direct-drive robot manipulators in bilateral master-slave system with environmental uncertainties and communication delay. A general condition based on the structured singular value μ for robustness of a bilateral manipulator is derived. The proposed control methodology can guarantee the robust stability and the robust performance for environmental uncertainty, perturbation of operator dynamics, perturbation of master and slave robot manipulator dynamics and constant communication delay of the master-slave system. Several experimental results show the effectiveness of our proposed approach for various environmental uncertainties and constant communication delay.

We have proposed a new approach to utilize the infinitesimal reflected waves by integrating the reflected waves. This method enables robot to move in unknown environments as it controls its speed smoothly. It is implemented using a transducer with a scanning system, and has a great ability of obtaining the traversable area for a robot in unknown environments in which the distance cannot be precisely measured. The proposed system, termed the integration-type ultrasonic wave sensor system, has following arbitrary parameters that the system designer can determine: sampling time of sonar, sampling time of scanning, range of scanning, speed of scanning, and so on. It is very difficult to determine them appropriately in response to changing environments or tasks. In this paper, we try to determine the arbitrary parameters of the integration-type ultrasonic wave sensor system on the basis of the 'consciousness'. We have proposed the speed control or obstacle avoidance by using the integration-type ultrasonic wave sensor system, and we define it as the unconsciousness movements. On the other hand, higher-level tasks such as navigation or map building are defined as the consciousness movements. The purpose of our research is to propose a simple construction and controlling method of the robot system by introducing the concept of consciousness. In preparation for this purpose, some experiments and discussion will be performed in this paper.

Recently, bicycles are widely used as a convenient transportation tool. But from a viewpoint of wide use for the future aging society, it is problem to pedal on rider's own. As well known, power assistance bicycle has already been used. The power assistance bicycle helps the elderly people or the people who has weak legs to expand their field. However, existing power assistance bicycle doesn't take running environment and rider's condition into account. The new control algorithm for power assistance bicycle is proposed in this paper. Human input and running friction are estimated as a disturbance torque with a disturbance observer. By using high pass filter (HPF), human input is separated from running friction. This method realizes power assistance bicycle without torque sensor. Disturbance observer compensates running friction. Compliance control is applied to make the bicycle have desired compliance. The effectiveness of this control algorithm is verified by numerical and experimental results.

An optomechatronic system based on one-dimensional ultrasound detector array for intravascular pressure measurement is presented. The proposed structure is based on optical linear ring resonators array. The applied pressure on ring resonator will change the physical parameters such as effective cross section of the ring and finally the effective index of refraction. The transmitted intensity changes and the measured output power could be used for measuring the pressure. Also, for rejection of common errors after photodetectors blocks, differential amplifiers can be utilized. For incident light coupling to the rings, the integrated 3-dB couplers are used. The effect of pressure on single ring is simulated and the effects of coupling coefficient, coupler loss and ring resonator diameter are investigated. This structure easily can be extended to two-dimensional cases. So, the proposed structure can be used for intravascular imaging including low noise and high integration and precision.

This paper presents a prototype rotary/linear dual-axis positioning system consisting of a θ-Z actuator and a rotary-linear angle sensor. In the system, an aluminum rotor (moving element) can be moved along and rotated about the axis (Z) of a ceramic cylinder (driving rod). The θ-Z actuator is composed of a Z-piezoelectric actuator (maximum stroke: 12 μm) for linear motion, two θ-piezoelectric actuators (maximum strokes: 9.1 μm) with an added weight for rotation, a driving rod and a rotor. The two θ-piezoelectric actuators with the added weight are attached to the driving rod via a clamping device made with steel. The inner face of the rotor is made contact to the driving rod with a certain friction force. The linear-axis positioning employs the smooth impact drive mechanism to achieve a large stroke by applying a periodic saw-toothed motion from the Z-piezoelectric actuator to the rotor via the driving rod. Sinusoidal motions are applied to the θ-piezoelectric actuators for rotary positioning, which is with a different mechanism form the smooth impact drive mechanism. The stroke of the prototype system along the Z-axis, which is limited by the length of the cylinder, is designed to be 10mm and there is no limitation in the rotary motion. The positioning resolution and maximum speed along the Z-direction are approximately a few nanometers and 2.4mm/sec, respectively. The maximum revolution speed is approximately 50 rpm. An optical surface encoder is also designed for precision positioning of the rotor.

In this paper, we develop a holonomic omni-directional mobile vehicle with step-climbing ability. This system realizes omni-directional motion on flat floor using special wheels and passes over the step in forward or backward direction using the passive linkage mechanism. Our key ideas are the following topics. First topic is new omnidirectional mobile mechanism which consists of special wheels and passive linkage mechanism. Second topic is new passive linkage mechanism which can change its body configuration easily when the vehicle passes over the step. Third topic is wheel control reference derivation based on the body configuration, which changes passively during step climbing for reducing wheel slippage. Last topic is wheel control method which keeps the rotation velocity coordination among the wheels for reducing wheel slippage and increasing the step-climbing performance. We verified the effectiveness of our proposed method by the computer simulations and experiments. Utilizing our proposed mechanism and control systems, the vehicle has both omnidirectional mobility and step overcoming function. Furthermore, our developing vehicle can pass over the 128[mm] height step using the wheel which radius is 66[mm]. Its performance is three times larger than one of general wheeled vehicle.

This paper presents controller design conditions for dynamic variable structure systems in terms of linear matrix inequalities (LMIs). In our previous paper, we proposed the dynamic variable structure system and derived its controller design conditions using switching fuzzy model-based control approach. However, the controller design conditions were given in terms of bilinear matrix inequalities (BMIs). In this paper, by introducing the augmented system which consists of the switching fuzzy model and a stable linear system, we derive new controller design conditions in terms of linear matrix inequalities (LMIs) for the dynamic variable structure systems. A simulation result shows the utility of this control approach.

Taking as an example the focal plane control system for the largest optical telescope being built in our country, the paper focuses on Universal Motion and Automation Controller (UMAC) based servo control system with high accuracy, and analyzes its design scheme. The scheme analysis and preliminary test demonstrate a broad outlook of UMAC based application in complex control systems with high precision, real time, fast action, easy adaptation and open architecture. A brief research summary and a preliminary test of the Focal Plane Control System (FPCS) are presented. The FPCS is one of components for the control system of Large sky Area Multi-Object fiber Spectroscopic Telescope (LAMOST), which is a national large scientific project. The design scheme features distributed, hierarchical and expansible network architecture with UMAC based control technology. A number of advanced techniques are integrated with some control software and hardware, which presents a solution to requirements of precision, real time and open architecture for the FPCS in the design of large optical astronomical telescopes.

This paper describes high-speed positioning of a surface motor-driven planar motion stage with a XYθ_z surface encoder. The surface motor consists of two pairs of linear motors. The magnetic array is mounted on the platen and the stator winding of the linear motor on the stage base. The platen can be moved in the X- and Y-directions by the X-linear motors and the Y-linear motors, respectively. It can also be rotated about the Z-axis by a moment about the Z-axis by the X- or Y-linear motors. The surface encoder consists of two two-dimensional angle sensors and an angle grid with two-dimensional sinusoidal waves on its surface. The angle grid is mounted on the platen. The sensors are placed inside the stage for a compact design of the stage system. The surface encoder is improved for high speed positioning. Measurement errors of the surface encoder using two kinds of detectors, the quadrant PD and two-dimensional PSD, are estimated by simulation. A modification of the motors for increasing the speed of the stage is also carried out. Verification experiments of the improved system are also performed.

This paper describes the replication of a precision sinusoidal grid surface, which is used as the measurement reference of a surface encoder for measurement of planar motions. The profile of the grid surface is a superposition of sinusoidal waves in the X-direction and the Y-direction with spatial wavelengths of a hundred micrometers and amplitudes of a hundred nanometers. The master surface is fabricated on a diamond turning machine equipped with a fast tool servo. Two kinds of replication methods, the hot embossing and UV casting are employed for replicating the grid surface on polymer materials. The replication on a glass plate is also carried out by UV-casting. The replication systems and some experimental results are presented.

This paper describes a micro-angle sensor based on laser autocollimation. The sensor consists of a light source, an objective lens and a positioning-sensing device. The position-sensing photodetector, which is a quadrant photodiode, is placed at the focal position of the objective lens. Differing from a conventional autocollimator, the angle sensor employs a laser diode as the light source. The laser beam is collimated to a thin parallel beam with a diameter of 1 mm so that the angle sensor can be used to detect the surface local slope of a specimen. The thin laser beam also makes it possible to use a small target mirror for measurement of stage angular error motions. Because the sensitivity of angle detection does not depend on the focal length of the objective lens by using a laser source, an objective lens with a short focal length is employed for realizing a compact sensor size. The prototype micro-angle sensor has a dimension of 26mm x 22mm x 12 mm. The resolution is better than 0.1 arc-seconds. Optical design and experimental results of confirming the performance of the sensor are presented.

This paper describes the optical design of a new type surface encoder, which consists of a 2D angle grid and a 2D slope sensor for 2D position detection. The sensitivity of the new type surface encoder is two times higher than that of a conventional surface encoder through employing a double pass. A wave optical model of the surface encoder is established for the optical design. Simulations based on the wave optical model are carried out to analyze the behavior of the diffraction pattern on the photo-detector of the slope sensor. Basic experiments are also carried out to confirm the feasibility of the double pass surface encoder.

A linear time-invariant model can be described either by a parametric model or by a nonparametric model. Nonparametric models, for which a priori information is not necessary, are basically the response of the dynamical system such as impulse response model and frequency models. Parametric models, such as transfer function models, can be easily described by a small number. In this paper, we will expand and generalize the orthogonal functions as basis functions for dynamical system representations. To this end, use is made of balanced realizations as inner transfer functions. The orthogonal functions can be considered as generalizations of, for example, the pulse functions, Laguerre functions, and Kautz functions, and give rise to an alternative series expansion of rational transfer functions. It is shown how we can exploit these generalized basis functions to increase the speed of convergence in a series expansion.