Weiyang Lin

Valid data based normalized cross-correlation (VDNCC) for topography identification

Abstract The Normalized Cross-Correlation (NCC) function is a widely used pattern-matching method. However, when the input data have a void area created by non-rectangular data or outliers, the accuracy of the standard NCC function may decrease. Especially when the regional mean values under the NCC window have a significant difference in the global mean value, the possible mis-matching may affect the identification results. In this paper, a valid data based NCC (VDNCC) algorithm is proposed for eliminating the effect of the void area. The new algorithm prevents void areas from being included in the calculation by introducing the valid data templates. VDNCC obtains higher NCC values and probabilities of correct matching in the experiments. In the ballistics identification tests, the results show that VDNCC can enhance the capacity of identification based on the NCC function as the core.

Visual Servoed Zebrafish Larva Heart Microinjection System

As a typical vertebrate model animal, the zebrafish has become a popular organism. For studying a specific disease mechanism, an efficient zebrafish microinjection technology is required. Compared with widely researched embryo injection technology, zebrafish larva injection has not been developed intensively. In particular, zebrafish larva heart injection is generally operated manually, which is laborious and time consuming. In this paper, we present a visual servo system of zebrafish larva heart microinjection, capable of visually determining larva heart position and controlling the microactuators. The larvas body is first rotated to be perpendicular to holding pipette; then, a robust visual detection algorithm against different larva gestures is presented to obtain its heart position in the horizontal plane. A rolling model is further constructed based on a database of larva body slice figures to estimate the heart height. Experimental results show that a satisfied success and survival rate can be obtained by presented methods, which is comparable to manual operation.

Two time-scale observer-based robust motion controller design and realization of a linear actuator:

In this paper, an observer-based sliding mode controller is proposed for a high-accuracy motion plant to suppress the disturbances and improve the tracking performance. In particular, a two time-scale separation technology, which can recover the disturbance state in a faster time scale, is utilized to compensate the disturbances and improve the system robustness. The parameter identification is carried out to obtain the model coefficients with a high fitting rate. Such an identified model can allow the engineers to tune the controller’s gains highly enough when the system suffers from the measurement noises. Instead of the traditional low-pass filter, a differentiator is introduced for the velocity signal prediction and its discrete-time version is provided to attenuate the noises effect. To verify the effectiveness of the proposed approach, an adaptive robust control law is compared with the proposed one in terms of dynamic positioning error, robustness and rapid signal tracking, and the superiority and advantages can be illustrated by the experimental results.

A novel robust algorithm for position and orientation detection based on cascaded deep neural network

Abstract Estimating position and orientation of the object by using machine vision is essential in industrial automation. The traditional canny operator and Hough transform edge detection algorithm is widely used, but its accuracy and real-time object recognition in complex backgrounds are very limited. Other algorithms such as SVM and BP network are usually inaccurate for regression issues. In this paper, the method of a cascade of convolution networks is proposed which results in high precision pose estimates. SSD is utilized to obtain the bounding box of the object to narrow down the recognition range. Convolution neural network is utilized to detect the orientation of the object. This method can extract weak features of the sample image. In generally, the proposed method possess a greatly improved accuracy and recognition rate compared with the traditional algorithm.

Visual Detection and Two-Dimensional Rotation Control in Zebrafish Larva Heart Microinjection

Zebrafish larva heart microinjection has emerged as an important technology that enables delivery of a wide variety of vehicles into hearts of small individuals. However, traditional manual microinjection is laborious and time consuming, and most existing automated injection systems can be costly or complicated to assemble. In this paper, a semiautomated and simple-structure vision servosystem of Zebrafish larva heart microinjection is presented. A robust vision recognition algorithm is developed to locate heart of anesthetized larvae with different postures. A rolling model that utilizes slice figures of larva body is constructed to estimate larva heart height. The two-dimensional rotation control that positions and orients larva by a motorized translational stage and a glass capillary is also proposed to meet the requirement of successful holding. Experimental results show that the developed visual detection method can achieve a success rate of 100% when larva heart is towards left or right and 90% in other postures. The rotation control approach also achieves a high success rate of 97% with rotation resolution of 0.5

Active fault tolerant control of buildings for seismic loads in finite frequency domain

{}^{\circ}

Graph-based SLAM in indoor environment using corner feature from laser sensor

. As a high-efficiency and high-cost-performance microinjection system, our system has a potential application in other organs injection of Zebrafish larvae.

Sliding-mode-based robust controller design for one channel in thrust vector system with electromechanical actuators ☆

Abstract This paper deals with the problem of vibration suppression for seismic-excited buildings in finite frequency domain, and the possible faults caused by actuators are also considered in controller׳s design to improve the stability of the system. Firstly, with the consideration of seismic wave׳s effect, a mathematical model of building structure system is established. Then the finite frequency theory is introduced to the process of controller design in order to reduce seismic-excited building vibration over a certain frequency band and enhance disturbance suppression performance. Furthermore, taking actuators׳ faults into account, the active fault tolerance control method is also added to the finite frequency domain controller to compensate for the faults effect and remain the performance of building system at an acceptable level when faults occur. Finally, simulation of a three-degree-of-freedom linear building structure under earthquake excitation is given to illustrate the effect of the proposed approach.

Positioning control for a linear actuator with nonlinear friction and input saturation using output-feedback control

SLAM (Simultaneous Localization And Mapping) is considered a fundamental problem for robots to become truly autonomous, and it is one of the most popular topic in the field of mobile robotics. When robot works in a unknown environment, it should estimate the current position relative to the environment and meanwhile estimate the environment. When both localization and mapping must be solved concurrently, the problem is called SLAM. SLAM can be implemented in many ways such the Particle Filter, Extended Kalman Filter and Graph-based solution. Currently, one of the most widely used algorithms to solve SLAM is Graph-based solution. In this paper we present a method for robot to calculate its accurate location in indoor environment using graph based optimization. We describe a way how to extract feature from laser range data and how to associate the features, and construct a robot pose graph when robot move in 2D environment. In the last of the paper, we present two simulated robot pose graph to compare the optimization result. The experimental results demonstrate our graph based optimization method is workable.

Position feedback dynamic surface control for pneumatic actuator position servo system

Abstract In this paper, a sliding-mode-based robust controller is proposed for the single channel thrust vector system (TVC) to suppress the disturbances and improve the tracking performance. Specifically, the dead-zone input–output relationship is analyzed to depict the mount gap in the mechanical shaft. The system mathematic representation including the mechanical and electrical sections, which suffers from the dead-zone nonlinearity, frictions and unstructured disturbance, is constructed. An adaptive-fuzzy-based observer is developed to estimate and compensate the disturbances because the fuel combustion dynamic and frictions in TVC are inevitable but difficult to obtain the precise dynamic state. Based on the nominal model, a robust controller is designed via the sliding-mode variable structure approach, which is derived in the sense of Lyapunov stability theorem. Instead of the traditional hitting law in the sliding mode controller, the chattering problem due to the discontinuous switch law is addressed by a continuous function. In the end, various illustrative examples are provided to demonstrate the effectiveness of the designed method.