Process diagnosis system and method for the diagnosis of processes and states in an technical process

• US 6,108,616 A
• Filed: 07/27/1998
• Issued: 08/22/2000
• Est. Priority Date: 07/25/1997
• Status: Expired due to Term

• Current Assignee: Ralf Otte
• Original Assignee: ABB Patent GmbH (ABB Limited)
• Inventors: Weisang, Claus, Borchers, Hans-Werner, Speh, Rainer, Otte, Ralf
• Primary Examiner(s): SHAH, KAMINI S
• Application Number:
  US09/123,079
• Time in Patent Office: 757 Days
• Field of Search: 702/183, 702/185, 702/182, 702/189, 702/57-59, 700/51, 700/286, 73/61.48, 706/10, 706/907, 714/26
• US Class Current: 702/183
• CPC Class Codes: G05B 2219/31211 Communicate diagnostic data from intelligent field device controller to central
• CPC Class Codes: G05B 2219/31434 Zone supervisor, collects error signals from, and diagnoses different zone
• CPC Class Codes: G05B 2219/33277 Distributed system with host as leader, host with multiple of agents
• CPC Class Codes: G05B 2219/33296 ANN for diagnostic, monitoring
• CPC Class Codes: G05B 23/0213 Modular or universal configuration of the monitoring system, e.g. monitoring system having modules that may be combined to build monitoring program; monitoring system that can be applied to legacy systems; adaptable monitoring system; using different communication protocols
• CPC Class Codes: G05B 23/0254 based on a quantitative model, e.g. mathematical relationships between inputs and outputs; functions: observer, Kalman filter, residual calculation, Neural Networks
• CPC Class Codes: G05B 23/0297 Reconfiguration of monitoring system, e.g. use of virtual sensors; change monitoring method as a response to monitoring results
• CPC Class Codes: Y10S 706/907 Power plant

Abstract

The invention relates to a process diagnosis system and a method for the diagnosis of processes and states in a technical process, in particular for the diagnosis of a power station process. The invention is a structured multi-agent system which corresponds to the process and has a plurality of autonomous diagnostic agents. The diagnostic agents in each case contain a neural network, with whose aid a reference behavior of process components that are to be monitored can be learned. In addition, automatic adaptation to a new reference behavior is also possible by the diagnostic agents.

Claims

1. A process diagnosis system for the diagnosis of processes and states of a technical process controlled by a process control system having a plurality of components, the process diagnosis system comprising:

a multi-agent system structured in accordance with a process and having a plurality of autonomous diagnostic agents distributed over a plurality of components of a process control system controlling the process;

said diagnostic agents each having a sensor module for the acquisition of process values and messages from the process control system, an actuator module for reacting on the process, a communications module for exchanging information between said diagnostic agents, an intentions module for storing task and objective programs of said diagnostic agents, and a processing module connected to said sensor module, said actuator module, said communications module and said intentions module, said processing module storing a process model, storing information about the process and diagnostic tasks and carrying out diagnostic functions; and

said processing module having an integrated neural network for learning a reference behavior of the components of the process being monitored and for enabling automatic adaptation to a new reference behavior.

2. The process diagnosis system according to claim 1, wherein some of said diagnostic agents are integrated into some of the components forming intelligent field instruments.

3. The process diagnosis system according to claim 1, wherein said neural network is a network for enabling adaptation of process changes wherein the process changes can be learned automatically without losing relevant process relationships that have already been learned.

4. The process diagnosis system according to claim 3, wherein said neural network is a locally approximating network.

5. The process diagnosis system according to claim 1, wherein said multi-agent system is for diagnosing processes and states of a power station process.

6. The process diagnosis system according to claim 1, wherein the components of the process control system include field instruments and said actuator module prescribes at least one of set points and setting parameters of the field instruments.

7. A method for carrying out a process diagnosis, which comprises:

providing a plurality of diagnostic agents to a process control system for monitoring process components of the process control system during a process, the diagnostic agents including lower order diagnostic agents and a higher order diagnostic agent forming a multi-agent system;

learning in a first commissioning phase reference behavior of the process components of the process to be monitored by each of the lower order diagnostic agents and storing the reference behavior as intended behavior;

determining in an application phase actual behavior of the process components and comparing the actual behavior with the stored intended behavior by each of the lower order diagnostic agents;

reporting any detected deviations in a form of deviation reports to the higher order diagnostic agent by each of the lower order diagnostic agents;

carrying out a diagnosis in the higher order diagnostic agent on a basis of the deviation reports received from the lower order diagnostic agents and on reference behavior stored by the higher order diagnostic agent; and

if no fault is detected, reporting with the higher order diagnostic agent to the lower order diagnostic agents that there is no fault present and that a reference model in at least one of the lower order diagnostic agents is to be adapted; and

if a fault is detected by the higher order diagnostic agent, reporting the fault with the higher order diagnostic agent to a superior higher order diagnostic agent thereof or outputting a fault report with a reporting device.

8. The method according to claim 7, which comprises automatically adapting each of the lower order diagnostic agents to a new reference behavior if the higher order diagnostic agent releases an adaptation request to each of the lower order diagnostic agents.