# US 2021/0231747 A1 · Partial discharge determination device and method

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

- Number: US 2021/0231747 A1
- Document ID: US20210231747A1
- Status: Published application
- Kind: A1
- Application number: 17/098,771
- Filed: 2020-11-16
- Published: 2021-07-29
- Earliest claimed: 2020-01-28
- Assignees: Hitachi, Ltd.
- Applicants: Hitachi, Ltd.
- Inventors: Hiromichi YAMADA; Mitsuyasu KIDO
- Web page: https://patentgrep.com/patent/US20210231747A1

## CPC

- G01R 31/1272: {of cable, line or wire insulation, e.g. using partial discharge measurements (locating faults in cables G01R31/083)}
- G01R 31/52: Testing for short-circuits, leakage current or ground faults
- G01R 31/14: Circuits therefor {, e.g. for generating test voltages, sensing circuits (G01R31/1209 - G01R31/1227 take precedence; for testing switches G01R31/327)}
- G01R 31/58: Testing of lines, cables or conductors (testing of electric windings G01R31/72)
- G06N 3/0499: Feedforward networks
- G06N 3/08: Learning methods
- G06N 3/09: Supervised learning
- G01R: MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES (indicating correct tuning of resonant circuits H03J3/12)
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS

## Abstract

An object of the present invention is to propose a partial discharge determination device and method that can highly reliably determine the degree of progress of partial discharge occurring in an underground power transmission cable without preparing plural standard patterns. A distribution pattern of a set of the charge amount and the occurrence phase angle of each partial discharge occurring in a period of one or more cycles of an applied voltage of a power transmission cable is sequentially generated, a first classification process in which the pattern is classified into any one of classes on the basis of a standard distribution pattern for each class and a second classification process in which the distribution pattern are executed, a classification result by the first classification process is evaluated on the basis of a classification result by the second classification process, and the degree of progress is determined.

## Claims

### Claim 1

A partial discharge determination device that determines the degree of progress of partial discharge occurring in a power transmission cable, the device comprising:
a distribution pattern generation unit that sequentially generates a distribution pattern of a set of the charge amount and the occurrence phase angle of each partial discharge occurring in a period of one or more cycles of an applied voltage of the power transmission cable;
a first classification unit that classifies the distribution pattern generated by the distribution pattern generation unit into any one of classes on the basis of a standard distribution pattern for each class;
a second classification unit that classifies the distribution pattern generated by the distribution pattern generation unit into any one of clusters by clustering;
an evaluation unit that evaluates a classification result of the first classification unit on the basis of a classification result of the second classification unit; and
a determination unit that determines the degree of progress of the partial discharge occurring in the power transmission cable on the basis of an evaluation result of the evaluation unit.

### Claim 2 (depends on 1)

The partial discharge determination device according to claim 1,
wherein the evaluation unit calculates the reliability of the classification result of the first classification unit on the basis of the classification result of the second classification unit, and evaluates the classification result of the first classification unit on the basis of the calculated reliability.

### Claim 3 (depends on 2)

The partial discharge determination device according to claim 2,
wherein the evaluation unit calculates the reliability having the classification result of the first classification unit for the distribution pattern as a class i (i=0, 1, 2 and so on) using the following formula,
[Formula 1]
Ri=Σk=1nriAk (1)
where the reliability of the classification result of the first classification unit is Ri, the number of clusters classified by the second classification unit so far is n, and the square of a ratio of the class having a classification result i of the first classification unit and the cluster having a classification result Ak of the second classification unit to the distribution patterns in which the classification result by the first classification unit and the classification result by the second classification unit are obtained so far is riAk.

### Claim 4 (depends on 1)

The partial discharge determination device according to claim 1,
wherein the first classification unit classifies the distribution pattern generated by the distribution pattern generation unit into any one of the classes using a neural network having learned the standard distribution pattern for each class, and
wherein a learning unit that learns the distribution pattern on the basis of the evaluation result of the evaluation unit to update the neural network is further provided.

### Claim 5

A partial discharge determination method executed by a partial discharge determination device that determines the degree of progress of partial discharge occurring in a power transmission cable, the method comprising:
a first step of sequentially generating a distribution pattern of a set of the charge amount and the occurrence phase angle of each partial discharge occurring in a period of one or more cycles of an applied voltage of the power transmission cable;
a second step of executing a first classification process in which the generated distribution pattern is classified into any one of classes on the basis of a standard distribution pattern for each class and a second classification process in which the distribution pattern generated by the distribution pattern generation unit is classified into any one of clusters by clustering;
a third step of evaluating a classification result by the first classification process on the basis of a classification result by the second classification process; and
a fourth step of determining the degree of progress of the partial discharge occurring in the power transmission cable on the basis of an evaluation result in the third step.

### Claim 6 (depends on 5)

The partial discharge determination method according to claim 5,
wherein in the third step, the reliability of the classification result by the first classification process is calculated on the basis of the classification result by the second classification process, and the classification result by the first classification process is evaluated on the basis of the calculated reliability.

### Claim 7 (depends on 6)

The partial discharge determination method according to claim 6,
wherein in the third step, the reliability having the classification result of the first classification unit for the distribution pattern as a class i (i=0, 1, 2 and so on) is calculated using the following formula,
[Formula 2]
Ri=Σk=1nriAk (2)
where the reliability of the classification result of the first classification unit is Ri, the number of clusters classified by the second classification unit so far is n, and the square of a ratio of the class having a classification result i of the first classification unit and the cluster having a classification result Ak of the second classification unit to the distribution patterns in which the classification result by the first classification unit and the classification result by the second classification unit are obtained so far is riAk.

### Claim 8 (depends on 5)

The partial discharge determination method according to claim 5,
wherein in the second step, the generated distribution pattern is classified into any one of the classes using a neural network having learned the standard distribution pattern for each class, and
wherein a learning step of learning the distribution pattern on the basis of the evaluation result in the third step to update the neural network is provided.
