Process for detecting complete sequences and failed sequences in a situation recognition sequence

• US 5,917,731 A
• Filed: 11/14/1997
• Issued: 06/29/1999
• Est. Priority Date: 07/29/1994
• Status: Expired due to Fees

• Current Assignee: Bull Sas (Atos SE)
• Original Assignee: Bull Sas (Atos SE)
• Inventors: Pares, Pascal, Grossetete, Christine, Ferenczi, Marc
• Primary Examiner(s): MOISE, EMMANUEL LIONEL
• Application Number:
  US08/970,618
• Time in Patent Office: 592 Days
• Field of Search: 364/515, 364/578, 395/500, 706/45, 706/52, 706/914, 706/920, 607/27
• US Class Current: 703/17
• CPC Class Codes: G05B 19/0428 Safety, monitoring G05B19/0423 takes precedence
• CPC Class Codes: G05B 2219/23255 Object oriented programming, OOP
• CPC Class Codes: G05B 2219/24081 Detect valid sequence of commands
• CPC Class Codes: Y10S 706/914 Process plant

Abstract

The invention relates to a process of detecting check sequences in a situation recognition system, composed of a list of events and time constraints. For each check sequence, a sequence of events and intervals between possible dates of occurrence of each of these events is created in memory. The process manages an internal clock differing from universal time, representing the current date of the system, updates the sequences of events for each situation upon each occurrence of a new event. This causes an updating of the current date, an updating for each sequence of events memorized of the intervals between possible dates of occurrences, detection of all the sequences of events for which one of these intervals is empty, creation of new sequences of events integrating this new event, and a detection of complete sequences.

Claims

1. A process for detecting failed sequences in a situation recognition system, including at least one situation, said situation comprising a list of events and time constraints linking these events, for which a sequence of events is created in a memory that contains a list of events expected in the situation, and intervals between possible dates of occurrence of each of these events, said intervals being initially equal to !-∞

, +∞

, said process managing an internal clock that differs from universal time, representing the current date of the system, said process comprising;

updating the sequences of events for each situation upon each occurrence of an event, said updating causing;

updating the current date to the date at which the event has occurred;

updating, for each sequence of events memorized, the intervals between possible dates of occurrence of events not yet recorded as a function of the time constraints and the new current date;

detecting all the sequences of events for which one of the intervals is empty, said sequences of events for which one of the intervals is empty being designated as failed sequences;

creating new sequences of events integrating the new event; and

detecting complete sequences.

2. The process of claim 1, wherein the memory occupied by a sequence of events is freed when the sequence of events is detected as being a failed sequence.

3. The process of claim 1, wherein the memory occupied by a sequence of events is freed when the sequence of events is detected as being a complete sequence.

4. The process of claim 1, wherein the creation of a new sequence of events, corresponding to the recognition of an expected event, causes callup of a function known as Recognition, associated with the situation that corresponds to said new sequence of events.

5. The process of claim 1, wherein the detection of a complete sequence of events causes callup of a function known as Complete, which is associated with the situation corresponding to said complete sequence of events.

6. The process of claim 5, wherein a function known as Failed is called up only when the failed sequence is not included in any other sequence of events.

7. The process of claim 6, wherein said date comprises a dead time date, wherein said dead time date is calculated as being the minimum of the upper limits of the intervals between possible dates of occurrence of events expected by each of the sequences of events.

8. The process of claim 7, wherein upon each occurrence of a new event, when the date of occurrence of the new event is later than the dead time date, the step of updating the current date to the date at which the new event occurred is replaced by a step of updating the current date to the dead time date, and the steps of updating the intervals, detecting the sequences of events for which one of the intervals is empty and calculating a dead time date are performed.

9. The process of claim 1, wherein the detection of a failed sequence causes callup of a function known as Failed, which is associated with the situation corresponding to said sequence of events.

10. The process of claim 1, wherein a positive delay value is defined for each event comprising a situation, said delay expressing, in the form of a duration, a possible delay in the signaling in the occurrence of this event.

11. The process of claim 1, further comprising a step of calculating a dead time date, a date closest to which at least one sequence of events will have been aborted when an event that is expected does not occur before said date.

12. The process of claim 1, wherein the definition of the situations comprises generating a time graph for each of the situations, said time graph enabling:

displaying, on a screen, the events and the time constraints linking them, with the aid of graphical objects and arcs, representing the events and the constraints, respectively, and having respective attributes, including the names, assigned to the events that they represent and the time constraint intervals; and

verifying consistency of routes formed by the time constraints between the various expected events.

13. The process of claim 12, wherein the step of generating said time graph includes generating a set of nodes and a set of arcs, wherein:

one event is associated with each node;

an arc represents the temporal relationship that links two nodes;

a set of leaving arcs leaving a first given node expresses a conjunction of events expected after the occurrence of an event of a first type corresponding to said first given node; and

a set of arcs entering a second given node expresses a conjunction of events preceding the occurrence of an event of a second type corresponding to said second given node.