Safe system provided with neural circuit

• US 5,586,220 A
• Filed: 05/18/1994
• Issued: 12/17/1996
• Est. Priority Date: 09/09/1991
• Status: Expired due to Fees

• Current Assignee: GTI Industrial Automation Incorporated
• Original Assignee: GTI Industrial Automation Incorporated
• Inventors: Spiker, Rolf T. E.
• Primary Examiner(s): Downs, Robert W.
• Assistant Examiner(s): Hafiz, Tariq R.
• Application Number:
  US08/204,297
• Time in Patent Office: 944 Days
• Field of Search: 123/159, 376/215, 395/20-25, 395/27, 382/155-159
• US Class Current: 706/23
• CPC Class Codes: G05B 13/027 using neural networks only
• CPC Class Codes: G05B 9/03 with multiple-channel loop, i.e. redundant control systems

Abstract

Disclosed is an electronic control circuit of the fail-safe emergency shut-down type which includes an input circuit, a first signal processing circuit, in which information is stored beforehand relating to extreme values input signals are permitted to reach, and an output circuit. The first signal processing circuit includes at least one neural network configured from very large numbers of neurons operating as integrators and in real-time which operate in parallel and which are mutually connected on a large scale. The networks are implemented in hardware and the extreme values are distributed over the neurons. When one of the extreme values is exceeded, the output circuit generates a shut-off signal and a component controlled thereby is placed in a safe state.

Claims

1. An electronic fail-safe emergency shut-down control circuit comprising an input circuit, a first signal processing circuit in which information is stored beforehand relating to extreme values which input signals are permitted to reach, and an output circuit which generates a shut-off signal when one of said extreme values in said first signal processing circuit is exceeded, whereby a component controlled by said control circuit is placed in a safe state, wherein

the first signal processing circuit includes at least one neural network configured from very large numbers of mutually connected neurons operating in real-time on a large scale in parallel as integrators, with said at least one neural network in said first signal processing circuit implemented in hardware and wherein said extreme values are distributed over the neurons, said neurons being voltaically separated from each other to prevent common cause failure;

further including a second signal processing circuit in which information is stored relating to extreme values of permitted changes in the time of the output signals, wherein when one of said extreme values in said second signal processing circuit is exceeded the output circuit generates an alarm signal which can control an alarm signalling device;

wherein the second signal processing circuit includes at least one neural network configures from very large numbers of mutually connected neurons operating in real-time in parallel as integrators, with said at least one neural network in said second signal processing circuit implemented in hardware and wherein said extreme values are distributed over the neurons; and

wherein the input circuit includes a digital signal input filter, the output of which is connected to a digital-to-digital converter, the output of the digital-to-digital converter being connected to the first signal processing circuit and to the second signal processing circuit, and a first analog signal input filter, the output of which is connected to a first on-line dynamic sensor validation unit, the output of the first validation unit being connected to a trip-amplifier, and the output of the trip-amplifier being connected to a first signal processing unit.

2. The electronic control circuit as claimed in claim 1, wherein at least one of the digital signal input filters and the first analog signal input filter are of the voltaically separated type.

3. The electronic control circuit as claimed in claim 1, wherein the input circuit includes a second analog input filter, the output of which is connected to a second on-line dynamic sensor validation unit, the output of the second validation unit being connected to an analog-to-digital converter, and the output of the analog-to-digital converter being connected to the second signal processing unit.

4. The electronic control circuit as claimed in claim 3, wherein the second analog filter is of the voltaically separated type.

5. The electronic control circuit as claimed in claim 3, wherein the second sensor validation unit is a neural network.

6. The electronic control circuit as claimed in claim 1, wherein the first sensor validation unit is a neural network.

7. The electronic control circuit as claimed in claim 1, wherein the trip-amplifier has an adjustable threshold value.