Progress Report. Mohammadtaghi G. Poshtmashhadi. Supervisor: Professor António M. Pascoal

Transcription

1 Progress Report Mohammadtaghi G. Poshtmashhadi Supervisor: Professor António M. Pascoal OceaNet meeting presentation April 2017

2 2 Work program Main Research Topic Autonomous Marine Vehicle Control and Navigation Systems for the inspection of Offshore Energy Infrastructures

3 3 Work program Core Theoretical Framework Navigation: Range only or range and bearing measurements provided by an Autonomous Surface Vehicle (ASV) Control: Cooperative Motion Control

4 4 Work program Applications (courtesy of PASS partners) Mid/close range offshore infrastructures inspection: Accurate navigation and path following of single and multiple AUVs for SURVEY operations. Close/very range offshore infrastructures inspection: Accurate navigation and sensor-based control of AUVs/ROVs for INTERVENTION and INSPECTION operations using acoustic systems (e.g. sonar)

5 5 Automatic Inspection Process Gathering data point from surrounding environment and constructing waypoints to follow. This can be done by profiling mechanical sonar. Path Generation Waypoint Guidance

6 Mechanical Profiling Sonar Micron Sonar Ultra Compact Imaging Sonar Specification Similar features to the SeaKing sonar in a compact housing 56 The Tritech Micron Sonar has set new standards in compact sonar technology. It is the smallest digital CHIRP sonar in the world. CHIRP technology dramatically improves the range resolution compared with conventional sonars it is a feature normally associated with much larger, more expensive systems. Ø56 If the new generation of very small and low cost ROVs are to develop their full potential it is essential they are equipped with the vital tools and sensors expected on larger vehicles. The Micron Sonar incorporates the very latest surface mounted digital electronics and many software features normally found only on full sized commercial systems. Based on experience gained from Tritech's world class range of SeaKing and SeaPrince sonars, the Micron incorporates the most advanced acoustic features and software available today. The sonar can be controlled by a customer supplied PC or laptop and it can be configured for either RS232 or RS485 protocols. The Micron has an auxiliary port to allow it to interface with other Tritech sensors. Benefits Extremely compact Simple to operate Cost effective and reliable Hard boot protected transducer Features 750m depth rating True acoustic zoom Digital CHIRP system RS232 and RS485 Target size measurement Applications Small ROV obstacle avoidance Target recognition AUV guidance Not to scale, dimensions in mm. Acoustic Operating frequency Beamwidth Maximum range Ø50 CHIRP centred on 700kHz 35 vertical, 3 horizontal 75m Minimum range 0.3m Range resolution approximately 7.5mm (minimum) Mechanical resolution 0.45, 0.9, 1.8 Scanned sector Variable up to 360 Continuous 360 scan? Yes Sector offset mode? Yes Electrical, Communications and Software Power requirement 12-48V DC at 4VA (average) Maximum cable length 1000m using RS485 Communication protocols RS485 (twisted pair), RS232 Surface control Computer using standard serial port, SeaHub or USB-RS232/RS485 converter Control software Tritech Seanet Pro, Micron software or low-level command protocol Software features True acoustic zoom, instant reversal, image measurement, inverted head operations Document: 0650-SOM-00004, Issue: 02 Physical Weight in air Weight in water Depth rating Temperature range 324g 180g 750m standard, 3000m optional -10 to 35 C (-20 to 50 C in storage) Specifications subject to change according to a policy of continual development. Document: 0650-SOM-00004, Issue: 02 Marketed by: Tritech International Ltd Peregrine Road, Westhill Business Park Westhill, Aberdeenshire, AB32 6JL United Kingdom +44(0)

7 7 Pure Pursuit guidance method waypoint Ocean current velocity Course angle Heading angle surge speed with respect to water

8 Pure Pursuit guidance method (continued) In Pure Pursuit guidance method, The vehicle aligns its velocity along the LOS vector. It can be done using a control law for angular velocity 8

9 Pure Pursuit guidance method (continued) 9 AUV kinematics and dynamics model Inner loop controller Constant Pure pursuit guidance algorithm Current velocity vector is unknown and should be estimated. Assumed to be constant

10 Pure Pursuit guidance method (continued) ρ is vehicle minimum turning radius. It is clear that when course angle χ is greater than the slope of LOS vector, the vehicle turns backward in order to point toward the waypoint. When the total velocity vector is facing in the opposite direction of the waypoint, constant rotation will be applied in order to keep it aligned with the waypoint. In case of no current the course angle is the same as yaw angle. 10

11 Obstacle Avoidance 11 Mechanical profiling sonar is used to detect obstacles online based on the range and bearing measurement

12 Obstacle Avoidance (Continued) 12 Based on obstacle position, the pre-planned path will be modified. Vector field histogram method (VFH) is used to determine the new steering direction for vehicle online. New planned path Obstacle Nominal path

13 13 VFH method for Obstacle Avoidance Borenstein, Johann, and Yoram Koren. "The vector field histogram-fast obstacle avoidance for mobile robots." IEEE Transactions on Robotics and Automation 7.3 (1991):

14 VFH method for Obstacle Avoidance (Continued) 14 VFH avoids obstacles using latest sensor readings Create local occupancy grid based on recent readings Polar histogram based on probability vs angle VFH+ includes kinematic limitation like min. turning radius

15 VFH method for Obstacle Avoidance (Continued) Path chosen by cost function G having three weighting parameters The VFH algorithm considers multiple steering directions based on current, previous, and target directions. Once the steering direction found, the heading rate and surge speed will be modified such that the vehicle does not collide with the obstacle 15

16 16 VFH implementation in Matlab VFH block uses the laser range readings to check if the target direction computed using the Pure Pursuit block is obstacle-free or not based on the laser scan data. If there are obstacles along the target direction, VFH block computes a steering direction that is closest to the target direction and is obstacle-free.

17 VFH implementation in Matlab 17

18 Path Following and Obstacle Avoidance in Simulink 18 Path Following with Obstacle Avoidance in Simulink IsNewSonarMsg Ranges Angles RobotPose Ranges Angles TargetDir du_{rd} Waypoints Inputs CurrentPose TargetDir u_{rd} u_{rd} dr_{d} r_{d} Adjust Velocities to Avoid Obstacles du_{rd} drd u_{rd} r_{d} Waypoints CloseToGoal Compute Velocity and Heading for Path following Rate Control r_{d} CloseToGoal Outputs Simulation Rate Control

19 Test Site 19 Testing Facilities EXPO 98 Site, Lisbon, PT

20 Specification for Sensor and 20 Vehicle Parameters for profiling sonar sensor: number of readings in 360 deg: 100 Max range: 50 m Sensor noise: 1 m Vehicle parameter: Surge speed with respect to water: 1 m/s Max angular rate: 30 deg/s Min turning radius: 5 m

21 21 Specification for VFH Parameters for VFH algorithm: Safety distance: 10 m Distance limits varies based on situation: lower bound 1 m and upper bound should be changed in order to prevent wobbling in corners Min turning radius: 5 m

22 22 Result of wall following No current

23 23 Result of wall following with current

24 24 Future plan o Extending wall following algorithm to application of inspection of an offshore structure o Capturing camera or multi-beam sonar images attached to the vehicle o Localization algorithm combined with cooperative motion control of multi agent systems