# US 2021/0004430 A1 · Data analysis device, power flow analysis device, and data analysis method

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

- Number: US 2021/0004430 A1
- Document ID: US20210004430A1
- Status: Published application
- Kind: A1
- Application number: 16/978,336
- Filed: 2018-05-08
- Published: 2021-01-07
- Earliest claimed: 2018-05-08
- Assignees: Hitachi, Ltd.
- Applicants: Hitachi, Ltd.
- Inventors: Sumito TOBE; Shinichi INAGE; Tooru AKATSU; Kouichi HARA
- Web page: https://patentgrep.com/patent/US20210004430A1

## CPC

- G06F 17/13: Differential equations (using digital differential analysers G06F7/64)
- G01R 21/06: by measuring current and voltage (G01R21/08 - G01R21/133 take precedence)
- G06N 3/126: Evolutionary algorithms, e.g. genetic algorithms or genetic programming
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)
- G01R: MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES (indicating correct tuning of resonant circuits H03J3/12)
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS

## Abstract

On the basis of time-series data, a term-specific time-series data creation unit creates time-series data for each term that includes a derivative term; a linear sum generation unit generates, for each time point in the created time-series data, a raw value for the linear sum of the products of each term times a coefficient; a coefficient determination unit determines a coefficient for each term such that the total value of raw values for linear sums within the time interval or space interval covered by the time-series data is no greater than a set value, or approaches zero; and a governing equation output unit outputs, as a governing equation for the time-series data, the linear sum of the products of each coefficient determined by the coefficient determination unit times each term.

## Claims

### Claim 1

A data analysis device comprising:
a term-based chronological data generation portion that, based on input chronological data, generates term-based chronological data including a differential term assumed to be a term of a dominant equation for the input chronological data;
a linear sum generation portion that generates a positive value of the linear sum of products of each term and a coefficient at each time of the input chronological data; and
a coefficient determination portion that determines the coefficient of each term so that the sum of positive values of the linear sum in one of a time segment and a space segment targeted by the chronological data becomes smaller than or equal to a predetermined value or approaches 0.

### Claim 2 (depends on 1)

The data analysis device according to claim 1, further comprising:
a dominant equation output portion that outputs, as a dominant equation for the input chronological data, the linear sum of products of the coefficient determined by the coefficient determination portion and each term.

### Claim 3 (depends on 1)

The data analysis device according to claim 1,
wherein the term includes one of a differential term with respect to time or space, a power term of time, space, or physical quantity, a functional term of time, space, or physical quantity, and a product term acquired by mutually multiplying the terms.

### Claim 4 (depends on 1)

The data analysis device according to claim 1,
wherein a positive value of the linear sum represents one of a square value and an absolute value of the linear sum.

### Claim 5 (depends on 1)

The data analysis device according to claim 1,
wherein the coefficient determination portion determines a coefficient of each term based on the Gauss-Seidel method.

### Claim 6 (depends on 1)

The data analysis device according to claim 1,
wherein the coefficient determination portion omits a term having a coefficient smaller than a predetermined value in comparison with coefficients of other terms.

### Claim 7 (depends on 6)

The data analysis device according to claim 6,
wherein the coefficient determination portion determines a coefficient of each term based on a genetic algorithm.

### Claim 8 (depends on 1)

The data analysis device according to claim 1, further comprising:
a dominant equation evaluation portion that evaluates the dominant equation based on a result of comparing a solution obtained by discretizing the dominant equation and the input chronological data.

### Claim 9

A power flow analysis device comprising:
a term-based chronological data generation portion that, based on input chronological data, generates term-based chronological data related to a load flow equation for a power system;
a linear sum generation portion that generates a positive value of the linear sum of products of each term and a coefficient at predetermined times of the input chronological data; and
a coefficient determination portion that determines the coefficient of each term so that the sum of positive values of the linear sum in one of a time segment and a space segment targeted by the chronological data becomes smaller than or equal to a predetermined value or approaches 0.

### Claim 10 (depends on 9)

The power flow analysis device according to claim 9,
wherein an impedance map of each line of the power system is generated based on a coefficient of each term related to the load flow equation.

### Claim 11

A data analysis method comprising:
determining a coefficient of each term of a differential equation so that the differential equation reproduces input chronological data; and
outputting the differential equation represented by the determined coefficient as a dominant equation for the input chronological data.

### Claim 12 (depends on 11)

The data analysis method according to claim 11 comprising:
generating term-based chronological data for the differential equation based on the input chronological data;
generating a positive value of the linear sum of products of each term and a coefficient of the differential equation at predetermined times of the input chronological data; and
determines the coefficient of each term so that the sum of positive values of the linear sum in one of a time segment and a space segment targeted by the chronological data becomes smaller than or equal to a predetermined value or approaches 0.

### Claim 13 (depends on 11)

The data analysis method according to claim 11,
wherein a term of the dominant equation includes one of a differential term with respect to time or space, a power term of time, space, or physical quantity, a functional term of time, space, or physical quantity, and a product term acquired by mutually multiplying the terms.
