Bradley D., Russell D.W. (eds.) Mechatronics in Action: Case Studies in Mechatronics

Подождите немного. Документ загружается.

116 D. Toal, E. Omerdic, J. Riordan, and S. Nolan

Fig. 7.13 Control Allocation Express VI

7.3.5 Autotuning of Low-level Controllers

Between successive ROV missions, it is likely that some onboard

instruments/sensors/equipment will be added/removed/replaced, leading to

changes in the dynamic properties of the ROV (mass, moments of inertia, drag

properties, etc.). Controllers optimally tuned for a particular vehicle configuration

will not give the optimal performance in the case of a change in configuration.

The autotuning of Low-Level Controllers (LLCs) is an advanced feature of the

control system, yielding optimal controller performance regardless of

configuration changes. It is recommended that the autotuning is performed at the

beginning of a mission. Autotuning algorithms described by Miskovic et al. [13]

have been expanded for six DOF controllers. The autotuning process is performed

for each LLC (surge, sway, heave, roll, pitch, yaw) separately and in the case of

the last four DOFs, involves the following steps:

1. generate self-oscillations;

2. wait for transient stage to be finished;

3. measure amplitude and period of steady-state oscillations;

4. find new values of controller gains using tuning rules.

A novel set of tuning rules for underwater applications has been developed

which provides the optimal performance of low-level controllers in the case of

Multi-mode Operations Marine Robotic Vehicle – a Mechatronics Case Study 117

configuration changes and presence of disturbances (waves and ocean currents).

Autotuning of the Heave LLC is shown in Figures 7.14 and 7.15. The ROV is

oscillating (moving up and down) around a set point of 30 m.

Fig. 7.14 VR display during heave autotuning

Fig. 7.15 Autotuning parameters and real-time data monitoring

7.3.6 High Frequency Sonar Enabling at Seabed Operation

When operating in near-seabed mode for video camera footage acquisition, DVL

and altimeter sonar become unreliable due to the minimum blanking range of the

instruments (typically 0.5–1.0 m) and this is at the point of most threat from

seabed hazards. To give reliable height off bottom readings in this critical near-

seabed zone, special optical fibre whisker sensors [14] and high frequency (1

MHz) wide beam angle (30°) sonar which give range readings down to 2 cm [15]

have been developed.

118 D. Toal, E. Omerdic, J. Riordan, and S. Nolan

7.3.7 Interchangeable Inshore and Deep Water Winch System

The system has an ROV umbilical winch with a small footprint and weight

suitable for deploying the ROV from inshore vessels of opportunity. Depending

on survey boat size, the ROV is either towed to site (small boat) or craned in and

out of the water (medium sized boat). By design, the inshore winch system can be

swapped out with a second deep water ROV winch system and the overall system

operation remains unaffected enabling ROV operational flexibility from inshore to

the edge of the continental shelf. The winch is of an all-electric design. The

umbilical on this winch is specified to include 2x20 A, 250 VDC circuits, 3 x 5 A

220 VAC circuits, and four single-mode optical fibre cores.

Optical fibre passes are unusual on inshore ROV winch systems. However,

they are included to enable seamless switching from inshore (winch) operation to

deep water (winch) operation. The deep-water ROV umbilical winch allows for

deploying the ROV from vessels of opportunity. The winch can accommodate

1,500 m umbilical length.

7.4 System Testing

The integrated payload systems and navigation positioning instruments have all

been integrated and operated on previous surveys on ROVs and surface-towed

pontoons. The vehicle control systems, low level controllers, fault tolerant control,

and instrument controllers have been developed, integrated and tested in the

Virtual Underwater Lab, an integral part of the MPPT Ring [16] which facilitates

hardware in the loop testing. The systems have further been tested in a test tank at

the University of Limerick.

Payload sensor systems are integrated and prepared for marine offshore

operations in the test tank which facilities a full power-on wet test of the

integrated systems prior to mobilisation. Testing of the complete integrated system

(vehicle and payload sensors) is planned for Foynes Dock on the Shannon estuary

prior to trials onboard the vessel Celtic Explorer.

7.5 Conclusions

This chapter presents the concurrent development of the University of Limerick

thrusted pontoon/ROV, a novel, multi-mode of operation marine robotics vehicle,

and MPPT Ring, a novel development platform featuring easy integration of

survey equipment, efficient planning and mission simulation, training of ROV

pilots, fault-tolerant control of ROVs, enhanced operator environment and survey

tools during mission execution and offline analysis of acquired data. The main

Multi-mode Operations Marine Robotic Vehicle – a Mechatronics Case Study 119

features, architecture and implementation issues are described in a compact way.

The MMRRC is building a flexible multi-mode of operation survey vehicle

(thrusted pontoon/ROV) for surface and underwater operations based on the

technologies and the architecture proposed in this chapter.

The vehicle has been designed with operational flexibility as a key criterion for

high-resolution near-seabed survey from shallow and inshore coastal waters out to

the continental shelf edge. The vehicle can be operated in surface deployment

mode either as a towed platform or as a thrusted pontoon. With the release of

buoyancy modules, the vehicle becomes neutrally buoyant and is operated as a

survey class ROV depth rated to 1,000 m.

Special features of the system include deployment system interoperability with

interchangeable winch and umbilical systems for both small inshore boats and

larger research vessel deployment in offshore operations. The vehicle has many

other advanced features to allow near-intervention/near-seabed operation,

including fault tolerant thruster control, novel high frequency sonar enabling at-

seabed operation and onboard electronics and computer control enabling real-time

reaction to environmental change/disturbance. The vehicle is also deployed,

operated and pilot controlled with a topside augmented reality system, giving a

pilot interface benefiting from virtual environment enhancement of the real world

operating scenario. The first sea test trials are expected to be performed in Winter

2008.

References

1. Grehan A, Wilson M, Guinan J, Riordan J, Molnar L, Omerdic E, Ullgren J, Le Guilloux E,

Toal D, Brown C (2005) ROV Investigations of Cold-Water Coral Habitats along the

Porcupine Bank Margin, West Coast of Ireland. 3rd Int. Symp. on Deep-Sea Corals, Miami

2. Grehan A, Toal D, Brown C (2006) ROV Investigations of Cold Water Coral Habitats in the

Porcupine/Rockall off the west coast of Ireland. The Irish Scientist Yearbook, 14:96

3. Molnar L, Omerdic E, Toal D (2005) Design of advanced AUV for deployment close to

seabed and other hazards. Ocean 2005 Europe, Brest

4. Toal D, Omerdic E, Nolan S, Al-Shamma'a A (2006) Intervention AUVs for permanent

seabed observatories, addressing some of the key technical challenges, 4th IEEE Int.

Workshop on Scientific Use of Submarine Cables & Related Technologies, Dublin

5. Omerdic E, Riordan J, Toal D (2008) MPPT Ring – Multi-Purpose Platform Technologies

for Subsea Operations. IFAC Workshop on Navigation, Guidance and Control of

Underwater Vehicles, Killaloe

6. Ridao P, Batlle E, Ribas D, Carreras M (2004) NEPTUNE: A HIL Simulator for Multiple

UUVs, Oceans'04

7. Omerdic E, Riordan J, Molnar L, Toal D (2006) Virtual Underwater Lab: Efficient Tool for

System Integration & UUV Control Development, 7th IFAC Conf Manoeuvring & Control

of Marine Craft, Lisbon

8. Riordan J, Toal D (2008) Performance Optimised Reverberation Modelling for Real-Time

Synthesis of Sidescan Sonar Imagery, IEEE J. of Oceanic Engineering, In print

120 D. Toal, E. Omerdic, J. Riordan, and S. Nolan

9. Wilson MFJ, O’Connell B, Brown C, Guinan JC, Grehan AJ (2007) Multiscale Terrain

Analysis of Multibeam Bathymetry Data for Habitat Mapping on the Continental Slope.

Marine Geodesy, 30:3–35

10. Riordan J (2006) Performance Optimised Reverberation Modelling for Real-Time Synthesis

of Sidescan Sonar Imagery. Ph.D. University of Limerick

11. Thurman E, Riordan J, Toal D (2007) Automated Optimisation of Simultaneous Multibeam

and Sidescan Sonar Seabed Mapping, Ocean 2007 Europe, Aberdeen

12. Omerdic E, Roberts GN (2006) Control Allocation for Over-Actuated Thruster-Propelled

Underwater Vehicles, in Advances in Unmanned Marine Vehicles, IEEE publication

13. Miskovic N, Vukic Z, Barisic M, Tovornik B (2006) Autotuning Autopilots for Micro

ROVs. Proceedings of 14th Mediterranean Conf. Control and Automation (MED06),

Ancona

14. Toal D, Flanagan C, Lyons WB, Nolan S, Lewis E (2005) Proximal object and hazard

detection for autonomous underwater vehicle with optical fibre sensors. Robotics and

Autonomous Systems, 53(3-4):214–229

15. Nolan S, Toal D (2008) A low directivity ultrasonic sensor for collision avoidance and

station keeping on inspection-class AUVs. IMarEST Journal of Marine Engineering, Science

and Technology, Part A, In print

16. Omerdic E, Riordan J, Toal D (2007) Easing Offshore Ops Through the Use of Augmented

Reality Support Tools, Unmanned Underwater Vehicle Showcase, Southampton

Chapter 8

Wireless Communication Technology

for Modular Mechatronic Controllers

Glen Bright

, Nkgatho S. Tlale

and Christopher M. Kumile

8.1 Introduction

Modular mechatronic systems are comprised of mechatronic subsystems that are

functionally complete and can be independently operated. These subsystems can

be readily fitted and connected to, or in combination with, additional subsystems

[1]. Advantages of the modular mechatronic design of systems include flexibility

to changing environmental needs, low cost due to standardisation of the

subsystems, interconnect ability and interchangeability of subsystems with each

other, and improved system reliability due to modular architecture [2].

Communication technology allows for the systematic integration of

mechatronic subsystems in order to achieve a complex integrated mechatronic

system which would be difficult to achieve without the use of information

technology (IT). Communication technology allows for the remote, real-time

control of mechatronic systems, e.g., the use of communication to send control

signals from haptic device and feedback signals from a robotic manipulator in

robotics-aided surgery where the surgeon and patient are in different locations

(e.g., countries).

In Figure 8.1, three different levels of communications are illustrated. The

choice of communication technology to be used is dependant on the complexity,

size and response time requirements of the tasks to be controlled. It is important to

match the time delay of the communication system with the time constant of the

system/process being controlled. If the time delays of the communication system

are not an issue, high level protocols can be used to control low level system

components/subsystems. Generally, high level system tasks can be adequately

controlled by use of a high level communication protocol (e.g., ethernet), while

University of KwaZulu-Natal, South Africa

Council for Scientific and Industrial Research, South Africa

Tshwane University of Technology, South Africa

122 G. Bright, N.S. Tlale, and C.M. Kumile

low level system tasks are adequately controlled with a low level communication

protocol (e.g., CanBus).

Communication technology in mechatronic systems also allows the

achievement of distributed control. However, this results in communication lags

due to the distributed architecture. This further introduces problems that are

concerned with timing, such as the lag effect associated with zero-order hold

(ZOH) and problems with respect to motion control. Problems of time variation

can be addressed in control design, e.g., by using robust control so that deviations

from nominal timing can be tolerated [3]. This chapter focuses more on the

modelling of such systems before the controller is implemented, i.e., the

determination of the dynamics of mechatronic systems in order to determine the

correct controller reaction time.

Controlled process(es)

System components

}

Mechatronics subsystems

Level 2 communications

Mechatronics system

Level 3 communications

Other systems

Decreasing time delay

requirement

Level 1 communications

Fig. 8.1 Classification of different mechatronic and communications levels

8.2 Modular Mechatronic Controllers

Mechatronic controllers are generally embedded controllers that use some form of

model or heuristic rules for the underlying system in order to achieve optimal

control of the system by monitoring sensor inputs and adjusting actuator outputs.

The physical and software components of a mechatronic controller are then; signal

conditioning hardware unit, signal processing software unit, central processing

unit (CPU), power electronics unit, communication circuits and communication

software unit.

Figure 8.2 shows a multiple-input multiple-output (MIMO) mechatronic system

with sensors, controller and actuators controlling a system/process. The system

can be described by its state variables x(t). The actuators outputs, u(t), are the

inputs of the system, while the outputs of the system, y(t), are the inputs to

sensors. A SISO system is comprised of only one sensor and one actuator.

Wireless Communication Technology for Modular Mechatronic Controllers 123

(t)

System

(t), x

(t), .... , X

(t)

Controller

m-buffersn-buffers

n-actuators

m-sensors

Fig. 8.2 Mechatronic controller on a process level

The use of communication technology requires communication nodes at

sensors, controllers and actuator nodes. The communication network is used to

transmit the information between sensors, controllers and actuators. The

transmission of information over a communication network takes some time,

which leads to communication lags. In many cases, the communication lags are

varying in a random fashion.

Communication lags depend on the configuration of the network and the

scheduling policy used. Factors like the network load, priorities of the other

ongoing communications, and electrical disturbances affect the communication

lags [4]. The communication lag at time interval k can be subdivided into three

parts:

• the communication lags between the sensor and the controller,

τ ;

• the computational lag in the controller,

τ ; and

• the communication lag between the controller and the actuator,

τ .

The control delay,

τ , can then be defined as the time from when a sensor

receives a signal to when it is used in the actuator as the control signal. Thus,

τ+τ+τ=τ . (8.1)

The Nyquist theorem states that the controller reaction time must be at least

half of the smallest time constant of the system to ensure proper system control

[5].

Depending on how the sensors, actuators, and controller nodes are

synchronised, several setups can be considered. Event-triggered controllers send

information as soon as it is available to the nodes (i.e., sensors and controllers).

The information is itself generated by some programmed activity (e.g., the signal

124 G. Bright, N.S. Tlale, and C.M. Kumile

being monitored achieving some specified level). Time-triggered systems transmit

the information based on some time model using the clock of the system as

reference. The node is able to start its activity at a pre-defined time (e.g., node’s

activities can be periodic). By implementing communication technology, a

distributed MIMO system can be achieved. Instead of centralising intelligence in

one controller, intelligence can be distributed on and between the different sensors

and actuators in the system.

Control and processing of data can only be done when data is available at the

node. Vacant sampling occurs when data does not arrive at the communication

node on time. To limit the level of vacant sampling, the buffers must be longer

than the worst case communication lag [6]. Buffering can be used between the

sensors, controllers, and actuators. This minimises the variation of the

communication lags between the communicating nodes and can be used to

produce constant communication lags. The disadvantage of using buffers is that

the control delay can become longer than necessary. This can lead to decreased

controller performance.

8.3 Communications Technology

Communications technology is used in mechatronic systems to achieve distributed

real-time control systems. The choice of the communication technology

implemented in a mechatronic system is dependent on the following:

• the spatial distances between the units of the system;

• the amount of information to be transferred via the communication technology

(or bandwidth); and

• the response time that is required from the communication technology.

For long distances (> 10 m, level 3 communication), communication protocols

such as the Transmission Control Protocol/Internet Protocol (TCP/IP), ProfiBus,

DeviceNet, FieldBus, Modbus, etc. are used [7]. These technologies are normally

used to connect one mechatronic system to a network of other mechatronic

systems. For medium distances (< 10 m, > 1 m, level 2 communications),

communication protocols such as the Controller Area Network (CAN), RS232,

RS485, RS422, GBIP, etc. are used. These protocols are also used in

instrumentation devices.

For short distances (< 1 m, level 1 communication), communication protocols

such as CAN, Inter-Integrated Circuit (I

C), Serial Peripheral Interface (SPI),

embedded TCP/IP, etc. are used. These protocols are normally used to connect

one microprocessor to others.

Wireless Communication Technology for Modular Mechatronic Controllers 125

8.4 Model-based Mechatronic Controllers

The difference between a model-based controller and a knowledge-based

controller is that the latter uses rules or heuristics, while the former uses some

model of the system in order to achieve control of the system.

For a SISO mechatronic system, let x(t) (= [x(t

), x(t

),…, x(t

)]) be a state variable

that can be used to fully describe a continuous-time varying system defined by f(t).

This infinite dimension system can then be described by Equation 8.2:

f(t) = g(x(t)). (8.2)

Let the controlled process be defined by:

)()(B)(A)( tttt vuxx ++=

(8.3)

where A is the state constant

B is the input constant

u(t) describes the input

and

v(t) is white noise with zero mean and has covariance R

The control loop of this mechatronic system can be formulated by sampling.

Assuming that the sampling period, h, is constant and greater than the delay from

the sensor to the actuator (i.e.,

ττh +> ). Integrating Equation 8.3 over a

sampling interval then results in Equation 8.4.

(

)

(

)

k1k

k0k1k

,, vuuxx +ττΓ+ττΓ+Φ=

−+

(8.4)

where

= x(kh) (8.5)

e=Φ

()

∫

τ−τ−

=ττΓ

Aca

dB,

(8.6)

()

∫

τ−τ−

=ττΓ

sca

dB,

(8.7)

and a variance of

Bradley D., Russell D.W. (eds.) Mechatronics in Action: Case Studies in Mechatronics - Applications and Education

Подождите немного. Документ загружается.