Xiang Jun Zou

Structure Design and Multi-Domain Modeling for a Picking Banana Manipulator

Based on agronomic and harvest characteristic analysis for bananas, according to bionic principle, a mechanism of picking banana manipulator simulating action of a person manually picking bananas was proposed, which was consisted of rotary mechanism, lifting mechanism, and shift mechanism, the body mechanical model of the picking banana manipulator was established. The multi-domain uniform model of a picking banana manipulator was developed employing the virtual simulation and modeling language Modelica. A body structure model for picking banana manipulator was established based on component models in DriveLib model library. Motion control’s mathematic models for picking banana manipulator were analyzed according to drive motor performance for the body structure, the simulation model of a mechanical motion control for the picking banana manipulator were realized in Dymola simulation platform. The validity models could provide a theoretical basis for picking banana manipulator primary design.

Modeling and Performance Analyzing of Helix Transmission Base on Modelica

Picking manipulator ontology structure used rotation and translational motion pairs to join the components together, and then a multi-joint motion mechanism was built which meet the picking space requirement. According to the characters of growth space of litchi fruit, here used helix transmission as the first axes of the picking manipulator, to achieve motion in vertical direction. This study introduces the application of multi-domain modeling and simulation when constructing and simulating a helix transmission. Use object-oriented technology, multi-domain modeling and simulation language Modelica to construct the motion system which integrates motor model, signal detection module, signal processing module and parametric screw transmission model in Dymola simulation platform. Test the helix transmission by simulation, using the mechanical structure, sensor technology, signal processing technology. Simulation results show that, the modelica can provide a valid simulation for complex electromechanical system and support for parametric design.

Research on Manipulator Positioning Based on Stereo Vision in Virtual Environment

Inherent uncertainties always exist in the positioning picking manipulators in complex environment, modeling and simulation of positioning. The manipulators were discussed based on binocular stereo vision in virtual environment (VE). Based on stereo vision, a method how virtual manipulators locate picking object by human-computer interaction (HCI) was proposed. The data input from vision were mapped to virtual picking manipulators so that it could enable the positioning and simulation with route and events-driven mechanism. The positioning experimental platform in VE consists of hardware of CCD stereo vision and simulation software. The visualized simulation system was exploited by EON SDK. The simulation of manipulator’s positioning was realized in VE by the platform. This method can be used for virtual robot to locate objects’ long-distance positioning in complex environment.

Picking Robot Camera Calibration System Based on OpenCV

Aim at the effect of the radial distortion and tangential distortion, the camera calibration principle, geometric model and complex equation solution were studied, in order to reduce the artificial participation in proofreading and error correction, then the camera’s software precise calibration module based on OpenCV was developed. The results showed that the system was quickly speed, high precision and could satisfy the requirements of robot vision navigation system and built the foundation for the next step research.

Application of Adaptive Fuzzy-PID Control in Servo Spindle Rigid Tapping

Rigid tapping of NC Tapping center is an electromechanical control system which is strongly nonlinear, uncertain and difficult to precise modeling. It is unavoidable to be affected by some nonlinear and uncertain factors such as friction and mechanical resonance. The traditional control method could not fundamentally solve the contradiction between rapid stability and high stable precision of the system response. As an intelligent control method, fuzzy control is actually a kind of PID regulator, and it exists a strong complementary to PID control in the system dynamic and static characteristics. This paper introduced fuzzy adaptive PID control applied to NC rigid tapping in engineering practice, the control method need not to build the controlled object, using mathematical model of fuzzy reasoning method for the online PID parameters auto-tuning, it met the system well for the rapidity and smoothness of comprehensive index requirement and technical realizability, finally through test platform test, numerical control machine tool actual tapping was validated.

Study on Location Simulation for Picking Manipulator in Virtual Environment Based on Information Fusion of Multi-Sensor

In picking manipulator location system, it is the key problem that the positions of obi-object and picking manipulator are exactly determined in complex environment. Based on multi-sensor information fusion method, a data fusion system of multi-sensor integrating laser-sensor for absolute location with ultrasonic-sensor for inspection impediment was presented. Firstly, data collection and fusion were implemented employing a two- level distribution system. Secondly, the method of data collection and fusion in virtual environment was discussed, and the result data could drive picking manipulator 3D model to dynamically move in real-time using event and route mechanisms provided by virtual environment, which could simulate the process of picking manipulator being accurately located. Finally, a location simulation system was developed by VC++ and EON SDK.

Research on the NC Woodworking Curve Band Saw CAM

To achieve the processing program and simulation verification on NC Woodworking Curve Band Saw (WCBS) that facilitates the industrialization and practicality of the woodworking machinery, the articles proposes the CAM system modules by the analysis of its composition structure and working principle of the machine and the building-up of its kinematics model. The article also explains the basic functions for each module and develops the CAM systems of MJS1325 WCBS with the use of VC++ and OpenGL. Reading and analyzing the CAD work piece graphics files, the CAM systems is able to generate the NC program and realize the simulation of machining in whole visual process environment including saws, fixtures, belt saw and work piece, etc. The CAM systems pass the prototype test and see a favorable effect.

Stereoscopic Positioning of Intelligent Fruit Harvesting Robot: a Review

Target stereoscopic positioning is a significant sign of intelligent robot. The paper first described the research status of harvesting robot target stereoscopic positioning, discussed senor and stereoscopic positioning applications. Then, method and algorithm of rapid identification in target image were stated. In addition, version positioning, mechanism positioning, comprehensive positioning, associated positioning error, random positioning systematic error and application and existing problem of robot were introduced in this paper. Finally, harvesting robot positioning error compensation was proposed with the aiming at realizing robot precision positioning and expecting a great prospect of intelligent harvesting robot.

The Correlative Positioning Error Compensation of the Vision System and the Robot Mechanism Based on BP Neural Network

Error analysis and compensation of positioning is one of the basic problems for robot control. There are some limitations to study the error compensation either from the perspective of vision system or mechanism of the robot. Therefore, this paper studied the error compensation from both aspects with BP neural network model. Firstly, aiming at the depth values obtained by the vision system, a depth error database was established and based on BP neural network the depth error was effectively compensated within 5mm in MATLAB simulation. Secondly, aiming at the mechanism of a robot prototype the error compensation simulation based on BP neural network in MATLAB was carried out, and the simulation result showed that the error was within ±1mm after compensation. Thirdly, the correlative positioning error was defined and its error compensation simulation was carried out in MATLAB with the BP neural networks of the vision system and the mechanism; the simulation results showed that the correlative positioning error was approximately zero in this study. Finally, the error compensation of the robot mechanism based on the BP neural network was carried out on the robot prototype HNrobot2, and the experiment results showed that the positioning error was by 60% less, which was within 5mm.

Modeling and Performance Simulation for a Picking Manipulator Based on Modelica

This paper presents a procedure and the results of structure design about a picking manipulator to be used for autonomous banana harvesting. Based on structure analysis and module partition for the picking manipulator, DriveLib model library was developed employing the virtual simulation and modeling language Modelica. A body structure model for the picking manipulator was established based on component models in DriveLib model library. Motion controls mathematic models for the picking manipulator were analyzed according to drive motor performance for the body structure, and then the partial differential equations were transformed into the ordinary differential equation easily solved by Modelica with the Laplace transform, mechanical motion simulation was realized. The validity models could provide a theoretical basis for picking banana manipulator primary design.