# US 2021/0241124 A1 · Identification of a characteristic of a physical system based on collaborative sensor networks > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 2021/0241124 A1 - Document ID: US20210241124A1 - Status: Published application - Kind: A1 - Application number: 17/165,714 - Filed: 2021-02-02 - Published: 2021-08-05 - Earliest claimed: 2020-02-03 - Assignees: Lawrence Livermore National Security, LLC - Applicants: Lawrence Livermore National Security, LLC - Inventors: Ryan Alan Goldhahn; Priyadip Ray; Braden C. Soper; Hao Chen; Deepak Rajan - Web page: https://patentgrep.com/patent/US20210241124A1 ## CPC - G06N 3/126: Evolutionary algorithms, e.g. genetic algorithms or genetic programming - G06N 3/086: using evolutionary algorithms, e.g. genetic algorithms or genetic programming - G06N 7/01: Probabilistic graphical models, e.g. probabilistic networks - G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS ## Abstract A node having a sensor and a computing device is provided for identifying a location of a source of a physical process. The node collects via the sensor a measurement of the physical process. The node repeatedly recalculates parameters until termination criterion is satisfied. The node receives parameters of a probability distribution and a gradient from other nodes. The node generates parameters based on the parameters and the gradients received from the other nodes. The node samples from a distribution of source locations based on the generated parameters. The node calculates a gradient derived from the sampled source locations, the generated parameters, and a joint probability of the sampled source locations and the measurement. The node sends to a subset of other nodes the generated parameters and the calculated gradient. When the termination criterion is satisfied, the generated parameters represent the probability distribution of the source location. ## Claims ### Claim 1 A method performed by a node associated with a sensor and a computing device for identifying a source location of a source of a physical quantity, the method comprising: collecting via the sensor a measurement of the physical quantity; and repeatedly re-calculating parameters until termination criterion is satisfied, for each of a plurality of other nodes, receiving parameters and a gradient, the parameters representing a probability distribution of the source location as determined by that node; generating parameters for the node based on the parameters and the gradients received from the other nodes; sampling source locations from a distribution of source locations based on the generated parameters; calculating a gradient derived from the sampled source locations, the generated parameters, and a joint probability of the sampled source locations and the measurement; and sending to a subset of other nodes the generated parameters and the calculated gradient wherein when the termination criterion is satisfied, the generated parameters collectively represent the probability distribution of the source location. ### Claim 2 (depends on 1) The method of claim 1 wherein the physical quantity is a gas that is released at the source location. ### Claim 3 (depends on 1) The method of claim 1 wherein the physical quantity is radiation that is released at the source location. ### Claim 4 (depends on 1) The method of claim 1 wherein the probability distribution is a normal distribution and the parameters are mean and variance of the normal distribution. ### Claim 5 (depends on 1) The method of claim 1 wherein the termination criterion is based on the calculated gradient. ### Claim 6 (depends on 1) The method of claim 1 wherein the termination criterion is based on a number of iterations of re-calculating the parameters. ### Claim 7 (depends on 1) The method of claim 1 wherein the calculating of the gradient is based on an Evidence Lower Bound relating to the probability distribution. ### Claim 8 (depends on 1) The method of claim 1 wherein the repeatedly recalculating the parameters is based on a Black Box Variational Inference algorithm. ### Claim at position 9 9-17. (canceled) ### Claim 18 A node for identifying a characteristic of an event based on a measurement relating to the event and based on probability distribution descriptions received from other nodes, each probability distribution description indicating a probability distribution of the characteristic, the node comprising: a sensor for collecting a measurement relating to the event; a communication component for communicating with other nodes; and a computing device that includes: one or more computer-readable storage mediums for storing computer-executable instructions for controlling the computing device to: receive via that communication component from each of a plurality of other nodes a probability distribution description and a gradient derived from samplings of the probability distributions by the other nodes; generate a probability distribution description based on the received probability distribution descriptions and gradients; sample values for the characteristic based on the generated probability distribution description; calculate a gradient derived from the sampled values, the generated probability distribution description, and a joint probability distribution of the sampled values and the measurement; and send via the communication component to a subset of other node the generated probability distribution description and the calculated gradient; and one or more processors for executing the computer-executable instructions stored in the one or more computer-readable storage mediums. ### Claim 19 (depends on 18) The node of claim 18 wherein the other nodes are a subset of nodes that are selected randomly. ### Claim 20 (depends on 18) The node of claim 18 wherein the gradient is calculated based on a Black Box Variational Inference algorithm. ### Claim at position 13 21-28. (canceled) ### Claim 29 (depends on 18) The node of claim 18 wherein the computer-executable instructions further include instructions to determine whether the generated probability distribution description and calculated gradient are informative and when not informative, suppress the sending of the generated probability distribution description and the calculated gradient. ### Claim 30 (depends on 18) The node of claim 18 wherein the computer-executable instructions further include instructions to receive from a sensor an indication to forward the calculated gradient to another node and forward the generated probability distribution description and calculated gradient to the other node. ### Claim 31 (depends on 18) The node of claim 18 wherein the computer-executable instructions further include instructions to prior to sending the generated probability distribution description and calculated gradient to other nodes, quantize the generated probability distribution description, the calculated gradient, or both to reduce bandwidth needed for the sending. ### Claim 32 (depends on 18) The node of claim 18 wherein the computer-executable instructions to determine whether the generated probability distribution description, the calculated gradient, or both are inconsistent with those received from other nodes, and when inconsistent, suppress the sending of the generated probability distribution description and the calculated gradient. ### Claim at position 18 -35. (canceled) ### Claim 36 A method performed by a node associated with a sensor and a computing device for identifying a characteristic of an event, the method comprising: collecting via the sensor a measurement of the characteristic; and repeatedly performing the following until termination criterion is satisfied, for each of a plurality of other nodes, receiving information relating to a value for the characteristic, the information including probability distribution information relating to a probability distribution of values for the characteristic as determined by that node; combining the information received from the other nodes to generate combined information that includes combined probability distribution information; sampling values of the characteristic from a probability distribution based on the combined probability distribution information; calculating information for the node that includes calculated probability distribution information derived from the sampled values and the combined probability distribution information; and sending to a subset of other nodes the calculated information wherein when the termination criterion is satisfied, the probability distribution of the combined information represents values of the characteristic of the event. ### Claim 37 (depends on 36) The method of claim 36 wherein the probability distribution information includes a gradient of a log likelihood of a measurement of a node at a last value of a Markov chain of sample values and includes a sampled value. ### Claim 38 (depends on 37) The method of claim 37 wherein the combining of the information includes calculating an average of the sample values and an average of the gradients of the log likelihood. ### Claim 39 (depends on 37) The method of claim 37 further comprising calculating a gradient of a log likelihood of the measurement and a gradient of a log of the probability distribution previously calculated by the node. ### Claim 40 (depends on 39) The method of claim 39 wherein the sampling of values is based on the gradient of the log of the probability distribution, the sampled values, and an average of the gradients of the log likelihood. ### Claim at position 24 41-49. (canceled)