# US 2017/0323039 A1 · Anomaly localization in a pipeline

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2017/0323039 A1
- Document ID: US20170323039A1
- Status: Published application
- Kind: A1
- Application number: 15/660,109
- Filed: 2017-07-26
- Published: 2017-11-09
- Earliest claimed: 2016-03-04
- Assignees: International Business Machines Corporation
- Applicants: International Business Machines Corporation
- Inventors: Tarun Kumar; Chandramouli Visweswariah; Bo Zhang; Rui Zhang
- Web page: https://patentgrep.com/patent/US20170323039A1

## CPC

- G06F 30/20: Design optimisation, verification or simulation (optimisation, verification or simulation of circuit designs G06F30/30)
- G06F 17/18: for evaluating statistical data {, e.g. average values, frequency distributions, probability functions, regression analysis (forecasting specially adapted for a specific administrative, business or logistic context G06Q10/04)}
- G06F 2111/10: Numerical modelling
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)

## Abstract

A method for locating an anomaly in a fluid transmission pipeline system is provided. The method may include receiving data for one or more physical conditions measured at an input and output of a pipeline portion, performing multiple simulations on a model of the pipeline portion to determine sets of simulated conditions that respectively correspond to simulated leak locations, determining a probability for a leak at one or more of the simulated leak locations by comparing one or more sets of simulated conditions to the received output data, and determining a highest probability location for the leak based on the probability for the leak at the one or more of the simulated leak locations. At least one simulation of the multiple simulations may be performed as a stochastic process and may be based on the received input data.

## Claims

### Claim 1

A method for locating an anomaly in a fluid transmission pipeline system, the method comprising:
receiving input data and output data for one or more physical conditions,
wherein the input data is measured at an input of a portion of a pipeline and the output data is measured at an output of the portion of the pipeline, and
wherein the one or more physical conditions comprise at least one of a flowrate, a pressure, or a temperature;
performing multiple simulations on a model of the portion of the pipeline, wherein each simulation determines a set of simulated conditions that corresponds to a simulated leak location, and wherein the multiple simulations determine a plurality of sets of simulated conditions that respectively correspond to a plurality of simulated leak locations;
determining a probability for a leak at one or more simulated leak locations from the plurality of simulated leak locations by comparing one or more sets from the plurality of sets of simulated conditions to the output data,
wherein determining the highest probability location for the leak is further based on stochastically updating the probability for the leak at the one or more simulated leak locations, and
wherein the one or more updated probabilities are based on stochastically selected input data; and
determining a highest probability location for the leak based on the probability for the leak at one or more simulated leak locations from the plurality of simulated leak locations,
wherein at least one simulation of the multiple simulations is performed as a stochastic process and is based on the input data.
