Patent
1994-05-18
1996-12-17
Downs, Robert W.
395 21, 395 24, G06E 100, G06E 300, G06F 1518
Patent
active
055862209
DESCRIPTION:
BRIEF SUMMARY
FIELD OF THE INVENTION
The invention relates to an electronic control circuit of the "Fail-Safe Emergency Shut-Down" type comprising an input circuit, a first signal processing circuit, in which information is stored beforehand relating to the extreme values input signals are permitted to reach, and an output circuit.
DESCRIPTION OF THE PRIOR ART
The expression "Fail-Safe" in the context of Emergency Shut-Down systems means in practice that the probability of a system defect as a result of which a non-defined final state occurs is as small as possible. The order of magnitude of this hazard P is equal to 10.sup.-6 or smaller. A system defect can occur due to all possible kinds of defects in components or combinations thereof.
A "Safety Related System" (SRS) of this type is used to determine any possible danger exceeding of one or more predetermined safety limits in a process outside the area that is controlled and regulated by the Process Control System and, after detection of this exceeding, to automatically return this process to a wholly safe state without an operator having to intervene and without an operator having to be informed of the cause.
The greatest possible reliability in combination with simplicity is obtained with a control circuit of the described type which has the feature that the said first signal processing circuit comprises at least one neural network configured from very large numbers of neurons operating as integrators and in real-time which operate in parallel and are mutually connected on a large scale, which networks are implemented in hardware and wherein the said extreme values are distributed over the neurons, wherein in the case when one of the said extreme values is exceeded the output circuit generates a shut-off signal whereby a component controlled thereby is placed in a safe state.
In a further embodiment the control circuit according to the invention is characterized by a second signal processing circuit comprising at least one neural network that is configured from very large numbers of neurons operating as integrators and in real-time which operate in parallel and are mutually connected on a large scale, which networks are implemented in hardware, in which second signal processing circuit information is stored relating to extreme values of the changes in the time that the output signals are permitted to have, wherein in the case when one of the said extreme values is exceeded the output circuit generates an alarm signal which can control an alarm signalling device.
This latter circuit is capable of detecting any potentially dangerous trend. After detection of a potential danger an alarm signal is transmitted to a monitoring station. On the basis of this alarm signal an operator can intervene in the system. If this intervention does not take place, the first signal processing circuit will, if the trend continues, detect after a period of time that one or more of the predetermined limits has been exceeded, whereby automatic shut-off of the process takes place without the intervention of an operator.
It is noted that a fail-safe disconnectable power supply is known from GB-A-2 150 374. The fail-safe principle is implemented by means of a transistor switch in combination with a fuse connected thereto in series and two relay contacts connected in parallel which can blow the fuse when the transistor circuit fails. This principle is known as the de-energize to blow principle.
A second fuse is further connected in series to the power supply circuit of the two parallel relays. This fuse is blown at each fail-safe shut-down command to prevent automatic restart and has to be replaced each time after a shut-down.
If regular testing of the two parallel relays is not possible, an inverter is then used to generate the fail-safe power.
Known from the Proceedings Tencon 87: 1987 IEE Region 10 Conference, part 3, Aug. 25, 1987, pages 893-898 "Intelligent control of variably configures systems using neural networks" is a rapid associative memory technique which is used for efficient
REFERENCES:
patent: 4515125 (1985-05-01), Buck et al.
patent: 4880596 (1989-11-01), Barrus et al.
patent: 5311562 (1994-05-01), Palusamy et al.
Proceedings Tencon 87: 1987 IEEE Region 10 Conference vol. 3, 25 Aug. 1987, Seoul, Korea.
Twenty-Second Asilomar Conference on Signals, Systems & Computers, vol. 1, 31 Oct. 1988, Pacific Grove, U.S.A.
INNC 90: International Neural Network Conference vol. 1, 9 Jul. 1990, Paris, France.
Downs Robert W.
GTI Industrial Automation B.V.
Hafiz Tariq R.
LandOfFree
Safe system provided with neural circuit does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Safe system provided with neural circuit, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Safe system provided with neural circuit will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-1998873