8. Прочитайте и переведите текст. Errorrecovery

Let us say that we are at the point where we have detected and classified an error. Now we have to start error recovery routine. How do we do this? There are two main points of view: rescheduling and active recovery. Rescheduling tries to avoid the negative effect of an error by rearranging the system's sequence of actions. It temporarily or permanently disregards the error. This is acceptable to some degree if the error was of stochastic nature, but this tactics does not improve the overall performance of the system, it does not add to the knowledge of the process and it does not consider the effects of recurring errors. It cannot be considered as a means of error recovery.

How does "active recovery" work? Active recovery signifies that the supervision system attempts to correct the error on-line with either manual or automatic salvage operations. Automatic error recovery is today possible to do at both the simple and the complex error levels, but becomes impossible for fatal errors.

Adaptive error recovery is a form of active recovery, theoretically more true to its name than any other form of recovery.

The most common way to use adaptive error recovery is to have smart sensors linked directly to the software.

The main difference between conventional error recovery and adaptability is that the automatic error recovery system works in a closed loop. Smart sensors and existing programs permit to correct the error. The classification scheme is simplified and the execution time can be, at least in theory, greatly reduced.

9. Ответьте на вопросы.

1. When do you start error recovery routine?

2. What is a rescheduling? What are its disadvantages?

3. What is an active recovery?

4. What kind of error may be recovered by automatic error recovery?

5. What is it necessary to use in adaptive error recovery?

6. What is the main difference between conventional error recovery and adaptability?

10. Подберите слова близкие по значению из данного ряда.

signify, repeat, start, mean, common, link, recur, connect, conventional, put into operation

11. Заполните пропуски нужными глаголами (to become, to reschedule, to signify, to recur, to attempt, to add) в соответствующей форме.

1. The system's sequence of actions…………

2. This…………nothing to our knowledge of the process.

3. These errors…………at very short intervals, they must be studied apart.

4. The supervision system…………to correct the errors on-line.

5. The unmanned machining…………a machining in an autonomic manufacture system without an operator.

6. After a break-down, that manufacturing process…………impossible.

12. Подберите термин (error, disturbance, failure, sensor), эквивалентный данному определению.

1. Instrument, which reacts to a certain physical condition by sending a signal (usually electrical by nature), transmitting information of this condition.

2. State of manufacture described by the quality index equal to zero or negative.

3. Special kind of failure caused by the control system or the person.

4. Anything which was not planned and influences the quality index of the process.

13. Из данного ряда слов выберите две группы из трёх синонимичных слов. Что они означают?

actuator, disturbance, error, sensor, failure, fault, malfunction, break-down, tolerance

14. Переведите текст письменно со словарем.

Research in the field of automatic process supervision and error recovery has been carried out over several years by various researchers. Automatic supervision and error recovery systems are software based and they utilize all types of transducers. The simplest way to design these programs is by using "hard programming" techniques. This means that the supervision and recovery routines are included into the robot programs at the robot controller level. This design is very complicated. If such an approach is taken, 80% of the program code will consist of sensor signals error recovery routines and "if-then" statements. The hard programming approach is time and code intensive and it does not solve the main problem: complex data structures. It also makes any debugging work extremely complicated. It would be better not to use the robot-controller-based software and write common programs at the cell-controller level.