# US 2025/0148394 A1 · Method and system for predicting regional gas consumption, and device and internet of things cloud platform

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

- Number: US 2025/0148394 A1
- Document ID: US20250148394A1
- Status: Published application
- Kind: A1
- Application number: 19/014,202
- Filed: 2025-01-08
- Published: 2025-05-08
- Earliest claimed: 2022-12-23
- Assignees: Zhejiang Cangnan Instrument Group Co., Ltd.
- Applicants: Zhejiang Cangnan Instrument Group Co., Ltd.
- Inventors: Zhongyang LI; Xiaoxiao SUN; Dongjin YU; Ming HUANG; Xin CHEN; Boyi HUANG; Chunyi YE
- Web page: https://patentgrep.com/patent/US20250148394A1

## CPC

- G06Q 10/06315: {Needs-based resource requirements planning or analysis}
- G06N 3/044: Recurrent networks, e.g. Hopfield networks
- G06N 3/0442: characterised by memory or gating, e.g. long short-term memory [LSTM] or gated recurrent units [GRU]
- G06N 3/045: Combinations of networks
- G06N 3/0455: Auto-encoder networks; Encoder-decoder networks
- G06N 3/08: Learning methods
- G06Q 10/04: Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem" (market predictions or forecasting for commercial activities G06Q30/0202)
- G06Q 10/0631: Resource planning, allocation, distributing or scheduling for enterprises or organisations
- G06Q 50/06: Energy or water supply
- G16Y 40/20: Analytics; Diagnosis
- G06Q: INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G16Y: INFORMATION AND COMMUNICATION TECHNOLOGY SPECIALLY ADAPTED FOR THE INTERNET OF THINGS [IoT]

## Abstract

A method and system for predicting regional gas consumption, and a device and a cloud platform. The future gas consumption of all gas users in an entire region is predicted according to a user-specified region and date; all networked gas users in the region are extracted during a prediction process by using spatial address matching technology; and then, by means of classifying the users and sampling the users in combination with statistics, representative sample users are extracted from all the users for respective prediction, thereby reducing the computation amount and computing resource consumption, which are required for model prediction. The method and system is mounted on a cloud platform, and the cloud platform receives, by means of Internet of Things technology, gas flow data that is uploaded by gas flow metering devices of all networked gas users, and also provides an interface for interacting with the users for the outside.

## Claims

### Claim 1

A method for predicting regional gas consumption, comprising:
S1: acquiring a user-specified target region, in which a gas consumption needs to be predicted, and a prediction date;
S2: according to target region, performing a spatial position matching on all gas users on basis of positions corresponding to all networked gas users, so as to obtain a set of the gas users which are located in the target region;
S3: on basis of category labels of all the networked gas users that are stored in a database in advance, obtaining, by means of querying, the category labels of each of the gas users from among the set of the gas users, the category labels of the gas users in the database are generated in advance by clustering gas consumption characteristics of all the networked gas users, the gas consumption characteristics are composed of a first gas consumption distribution sequence on weekdays and a second gas consumption distribution sequence on holidays of the gas users, and all newly added networked gas users need to be assigned corresponding category labels through clustering in the database;
S4: dividing all the gas users in the set of the gas users into user subsets according to category labels, wherein the gas users from among each of the user subsets have a same category label; and then, performing sampling on the gas users from among each of the user subsets, so as to accordingly obtain, for each of the user subsets, sample user sets which is capable of representing entire user subset;
S5: for each of the sample user sets, according to the category labels corresponding thereto, calling a gas consumption prediction model, which is constructed and trained for this type of the gas users, so as to obtain a gas consumption prediction value of each of sample users from among the sample user sets on the prediction date, wherein the gas consumption prediction model is a Seq2Seq model based on a LSTM recurrent neural network model, its input is a time information encoding day sequence for multiple consecutive days before the prediction date, and its output is the gas consumption prediction value on the prediction date, wherein the time information encoding day sequence of any day is formed by fusion of a gas consumption distribution sequence of that day, first encoding information of which day of the week that day belongs to, and second encoding information of whether that day is a holiday; and
S6: averaging gas consumption prediction values of all the sample users from among each of the sample user sets on the prediction date, multiplying a mean value by total number of gas users from among the user subsets that corresponds to the sample user sets, so as to obtain a total gas consumption prediction value of all gas users from among each of the user subsets, summing up total gas consumption prediction values of all of the user subsets, and obtaining, by means of prediction, a total gas consumption of all gas users in the target region on the prediction date.

### Claim 2 (depends on 1)

The method for the predicting regional gas consumption according to claim 1, wherein in the S1, the target region and the prediction date are input or specified through a Graphical User Interface (GUI) capable of human-computer interaction, and the target region is designated by an administrative division or by free selection on a map interface.

### Claim 3 (depends on 1)

The method for the predicting regional gas consumption according to claim 1, wherein in the S2, when matching the set of gas users located in the target region, spatial topological calculation is required to be performed on point features corresponding to positions of all the gas users and a surface feature corresponding to the target region, and the point features intersecting with the surface feature are the gas users located in the target region.

### Claim 4 (depends on 1)

The method for the predicting regional gas consumption according to claim 1, wherein in the S3, the clustering is implemented using a K-means clustering algorithm, gas consumption distribution sequences of each of the gas users on all non-holidays are averaged and the gas consumption distribution sequences of each of the gas users on all holidays are averaged before the clustering, two average sequences are concatenated to serve as a distance characteristic of the gas users and are used to calculate a distance value between the gas users and a clustering center during a clustering process, so as to classify the gas users into a cluster with a closest distance.

### Claim 5 (depends on 1)

The method for the predicting regional gas consumption according to claim 1, wherein a gas consumption distribution sequence is obtained by processing original data uploaded to a server by a gas flow metering device based on an Internet of Things technology, preferably, a time length of the gas consumption distribution sequence is 1 day, a sampling interval is 30 minutes to 2 hours, and each of sampling values represents a gas consumption of the gas users during the sampling interval.

### Claim 6 (depends on 1)

The method for the predicting regional gas consumption according to claim 1, wherein the gas consumption prediction model adopts a Seq2Seq model with an encoder-decoder structure, wherein an encoder part adopts a multi-layer first LSTM recurrent neural network model, the decoder part adopts a single-layer second LSTM recurrent neural network model, an attention layer cascaded between an encoder and an decoder is provided, a fully connected layer is cascaded after the decoder for output, input of first LSTM recurrent neural network model is the time information encoding day sequence for the multiple consecutive days before the prediction date, its a model output and a hidden layer vector are inputted into the attention layer to generate an attention weight through an attention mechanism, the model output of the first LSTM recurrent neural network model is weighted using the attention weight to obtain a context vector, the context vector is concatenated with the hidden layer vector of the first LSTM recurrent neural network model and inputted into second LSTM recurrent neural network model, and output of the second LSTM recurrent neural network model is concatenated with the context vector and inputted into the fully connected layer to obtain the gas consumption prediction value on the prediction date as final output of the gas consumption prediction model.

### Claim 7 (depends on 1)

The method for the predicting regional gas consumption according to claim 1, wherein sampling method is random sampling, stratified sampling, overall sampling, or systematic sampling.

### Claim 8 (depends on 1)

A computer electronic device, comprising a memory and a processor, wherein
the memory is configured to store a computer program, and
the processor is configured to implement a method for the predicting regional gas consumption according claim 1 when executing the computer program.

### Claim 9 (depends on 1)

An Internet of Things cloud platform, wherein the cloud platform receives, by means of Internet of Things technology, gas flow data uploaded by gas flow metering devices of all the networked gas users and also provides an interface for interacting with the users for the outside, and the cloud platform stores a computer program and executes a method for the predicting regional gas consumption according to claim 1 when the computer program is executed by a processor.

### Claim 10 (depends on 4)

The method for the predicting regional gas consumption according to claim 4, wherein a gas consumption distribution sequence is obtained by processing original data uploaded to a server by a gas flow metering device based on an Internet of Things technology, preferably, a time length of the gas consumption distribution sequence is 1 day, a sampling interval is 30 minutes to 2 hours, and each of sampling values represents a gas consumption of the gas users during the sampling interval.

### Claim 11 (depends on 2)

A computer electronic device, characterized in comprising a memory and a processor, wherein
the memory is configured to store a computer program, and
the processor is configured to implement a method for the predicting regional gas consumption according to claim 2 when executing the computer program.

### Claim 12 (depends on 3)

A computer electronic device, characterized in comprising a memory and a processor, wherein
the memory is configured to store a computer program, and
the processor is configured to implement a method for the predicting regional gas consumption according to claim 3 when executing the computer program.

### Claim 13 (depends on 4)

A computer electronic device, characterized in comprising a memory and a processor, wherein
the memory is configured to store a computer program, and
the processor is configured to implement a method for the predicting regional gas consumption according to claim 4 when executing the computer program.

### Claim 14 (depends on 5)

A computer electronic device, characterized in comprising a memory and a processor, wherein
the memory is configured to store a computer program, and
the processor is configured to implement a method for the predicting regional gas consumption according to claim 5 when executing the computer program.

### Claim 15 (depends on 6)

A computer electronic device, characterized in comprising a memory and a processor, wherein
the memory is configured to store a computer program, and
the processor is configured to implement a method for the predicting regional gas consumption according to claim 6 when executing the computer program.

### Claim 16 (depends on 7)

A computer electronic device, characterized in comprising a memory and a processor, wherein
the memory is configured to store a computer program, and
the processor is configured to implement a method for the predicting regional gas consumption according to claim 7 when executing the computer program.
