# A Smart Online Over-Voltage Monitoring and Identification System

^{1}

^{2}

^{*}

## Abstract

**:**

## 1. Introduction

## 2. Over-Voltage Monitoring System

#### 2.1. Over-Voltage Sensor

_{S}is the divider capacitance, R

_{S}is the divider resistance, R

_{P}is a matching resistance and P

_{S}is a protective unit. The voltage-dividing unit is composed of a dividing capacitance and dividing resistance, respectively constituting the high-frequency response and low-frequency response closed circuit. In order to reduce the magnetic coupling between the elements of the voltage-dividing unit and the output circuit, as well as to minimize the influence of residual inductance on the response characteristics of the sensor, the sensor structure adopts a coaxial cylinder structure. The matched resistance is connected to the signal transmission cable interface through the center of the sensor. The resistance and capacitance voltage-dividing elements are connected in parallel and arranged symmetrically around the circumference toward the center. The outermost shell of the whole sensor is the metal shell. It is designed to realize electromagnetic shielding for core components through the grounded metal shell.

#### 2.2. Online Monitoring System

## 3. Over-Voltage Identification System

- Online over-voltage identification.
- Trigger the suppression system at the next level, and alert the operator through human interface.
- Searching and analysis of past over-voltages from the database.
- Statistics of appearance frequency for different kinds of over-voltage in different power grids.

## 4. S-Transform and Feature Extraction for Over-Voltage Records

#### 4.1. S-Transform

#### 4.2. Feature of Over-Voltage

- (1)
- When a single phase-to-ground occurs, the voltage in the fault phase drops to zero, and the other two phase voltages increase. Typical waveforms are shown in Figure 5a.
- (2)
- When a fundamental ferroresonance takes place, a power frequency over-voltage is shown in the power system, the amplitude in three phases can be expressed by three cases:
- (i)
- The voltage in one phase drops but not equal to zero, and the other two increase (Figure 5b).
- (ii)
- The voltage in two phases drop and the other one increase.
- (iii)
- The voltages in three phases all increase.

**Figure 5.**(

**a**) Single phase-to-ground voltage; (

**b**) Fundamental ferroresonance; (

**c**) Mean amplitude curve of single phase-to-ground voltage; (

**d**) Mean amplitude curve of fundamental ferroresonance.

- (3)
- If a high frequency ferroresonance occurs, the voltage in the three phases contain 150 Hz, 250 Hz, 350 Hz harmonics (3, 5, 7 times of 50 Hz power frequency). A typical waveform is shown in Figure 6a.
- (4)
- If a sub-frequency ferroresonance occurs, the voltage signals in the three phases contain 25 Hz or 16.6 Hz harmonics (1/2, 1/3 of power frequency). A typical waveform is shown in Figure 6b.

**Figure 6.**(

**a**) High frequency ferroresonance; (

**b**) Sub-frequency ferroresonance; (

**c**) Mean amplitude curve of high frequency ferroresonance; (

**d**) Mean amplitude curve of sub-frequency ferroresonance.

- (5)
- When the arc ground fault takes place in the system, the arc at the fault point will extinguish and rekindle repeatedly. It is presented as repeated high-frequency oscillations on the waveform. A typical waveform is shown in Figure 7a.
- (6)
- When a capacitor switching occurs, the power voltage component does not change too much, and the high frequency oscillation is highly similar with the line switching. A damping harmonic with the frequency around 200 Hz will occur at the switching moment, which is the major feature of a switch capacitor. A typical waveform is shown in Figure 7b.

**Figure 7.**(

**a**) Arc grounding; (

**b**) Capacitor switching; (

**c**) Mean amplitude curve of arc grounding; (

**d**) Mean amplitude curve of capacitor switching.

- (7)
- The main feature of lightning and line switching over-voltage is shown in their high frequency oscillation. General, the power frequency component will not change too much after these two over-voltages take place. Typical waveforms of lightning and line switching are shown in Figure 8a,b.

**Figure 8.**(

**a**) Lightning; (

**b**) Line switching; (

**c**) Transient oscillation of lightning; (

**d**) Transient oscillation of line switching; (

**e**) Mean amplitude curve of lightning; (

**f**) Mean amplitude curve of line switching.

#### 4.3. Feature Extraction

_{nml}is the normal power frequency component’s amplitude; in this paper ${V}_{nml}=10.5/\sqrt{6}\approx 4.2866$.

Type | CQ1 | CQ2 | CQ3 | CQ4 | CQ5 |
---|---|---|---|---|---|

Capacitor switching | 0.6192 | 1.0220 | 0.9820 | 24.3205 | 185 |

Arc grounding | 0.3977 | 1.8304 | 0.5030 | 6.5734 | 190 |

High frequency ferroresonance | 0.4007 | 2.9556 | 1.6254 | 2.3694 | 150 |

Fundamental ferroresonance | 0.6145 | 1.6446 | 0.7443 | 2.0455 | 0 |

Sub-frequency ferroresonance | 0.4277 | 1.0160 | 1.5705 | 2.5649 | 25 |

Single phase-to-ground | 0.7110 | 1.7279 | 0.0125 | 2.3310 | 0 |

Line switching | 0.6757 | 1.0092 | 0.9891 | 15.7937 | 0 |

Lightning | 0.6280 | 1.0447 | 0.9654 | 43.1868 | 0 |

Type | CQ6 | ||||
---|---|---|---|---|---|

Line switching | 0.0098 | 0.0134 | 0.0303 | 0.0250 | 0.9215 |

Lightning | 0.3266 | 0.3437 | 0.2271 | 0.0290 | 0.0739 |

## 5. Over-Voltage Identification System

#### 5.1. Fuzzy Expert System

_{A}(x) presents how much x belongs to A, expressed as:

_{1}is A

_{1}and x

_{2}is A

_{2}… x

_{n}is A

_{n}then y is B

_{i}(I = 1,2,…,n) and B are linguistic values. There are many mature function expressions available for forming the membership function, such as triangular MF, trapezoidal MF, generalized bell MF and Gaussian MF. In this paper, so the distribution of characteristic quantity’s values are more closer to a normal distribution, so the Gaussian MF (normal distribution function) is employed as the membership function:

#### 5.2. Membership Function

_{s}= 1/NT, in this paper f

_{s}= 5 Hz), and two center points are designated as 15 Hz and 25 Hz.

_{D}

_{2}has two standard deviation σ, because the distribution of CQ4 are uneven on two sides of the mean value c.

MF | σ | MF | σ |
---|---|---|---|

μ_{A}_{1} | 0.0709 | μ_{D1} | 2.3009 |

μ_{A}_{2} | 0.0797 | μ_{D}_{2} | 7.6125/19.3283 |

μ_{B1} | 0.1478 | μ_{D3} | 7.9675 |

μ_{B2} | 0.1798 | μ_{E1} | 5 |

μ_{C1} | 0.1721 | μ_{E2} | 5 |

μ_{C}_{2} | 0.3578 | μ_{E3} | 15 |

μ_{C3} | 0.3405 | μ_{E4} | 5 |

#### 5.3. Fuzzy Rule Base

- Rule 1: If CQ1 is A2 and CQ2 is B1 and CQ3 is C3 and CQ4 is D3 and CQ5 is E3 then T1 with CF1.
- Rule 2: If CQ1 is A1 and CQ3 is C3 and CQ4 is D2 then T2 with CF2.
- Rule 3: If CQ1 is A1 and CQ3 is C3 and CQ4 is D1 and CQ5 is E2 then T3 with CF3.
- Rule 4: If CQ1 is A2 and CQ2 is B2 and CQ3 is C2 and CQ4 is D1 and CQ5 is E4 then T4 with CF4.
- Rule 5: If CQ1 is A1 and CQ3 is C3 and CQ4 is D1 and CQ5 is E1 then T5 with CF5.
- Rule 6: If CQ1 is A2 and CQ2 is B2 and CQ3 is C1 and CQ4 is D1 and CQ5 is E4 then T6 with CF6.
- Rule 7: If CQ1 is A2 and CQ2 is B1 and CQ3 is C3 and CQ4 is D3 and CQ5 is E4 then T7 with CF7.

_{A}

_{2}(CQ1), μ

_{B}

_{1}(CQ2), μ

_{C}

_{3}(CQ3), μ

_{D}

_{3}(CQ4), μ

_{E}

_{3}(CQ5)}

Type | CF1 | CF2 | CF3 | CF4 | CF5 | CF6 | CF7 |
---|---|---|---|---|---|---|---|

Capacitor switching | 0.8985 | 0.0178 | 0 | 0 | 0 | 0 | 0 |

Arc grounding | 0 | 0.7376 | 0 | 0.0074 | 0 | 0 | 0 |

High frequency ferroresonance | 0 | 0.0172 | 1 | 0.0083 | 0 | 0 | 0 |

Fundamental ferroresonance | 0 | 0.0214 | 0 | 0.9188 | 0.0111 | 0.0008 | 0.0001 |

Sub-frequency ferroresonance | 0 | 0.0263 | 0 | 0.0012 | 0.9905 | 0 | 0.0111 |

Single phase-to-ground | 0 | 0.0002 | 0 | 0.2535 | 0.0002 | 1 | 0 |

Line switching | 0 | 0.0013 | 0 | 0 | 0 | 0 | 0.8706 |

Lightning | 0 | 0.0124 | 0 | 0 | 0 | 0 | 0.8775 |

#### 5.4. Identification System Based on FES and SVM

Type | Characteristic Quantities | Identification Results | ||||
---|---|---|---|---|---|---|

CQ1 | CQ2 | CQ3 | CQ4 | CQ5 | ||

Capacitor switching | 0.5951 | 1.0143 | 1.0101 | 33.9705 | 185 | T1(Correct) |

0.5519 | 1.0415 | 1.0013 | 27.1107 | 190 | T1(Correct) | |

0.6330 | 1.1083 | 1.0194 | 45.2598 | 180 | T1(Correct) | |

Arc grounding | 0.2609 | 1.8658 | 0.4890 | 8.1343 | 145 | T2(Correct) |

0.1856 | 1.8265 | 0.7340 | 7.0991 | 150 | T2(Correct) | |

0.3136 | 1.4472 | 0.9158 | 17.1156 | 20 | T2(Correct) | |

High frequency ferroresonance | 0.3608 | 2.9598 | 1.5899 | 2.2907 | 150 | T3(Correct) |

0.3496 | 2.9571 | 1.6014 | 2.3197 | 150 | T3(Correct) | |

0.3370 | 2.9546 | 1.6147 | 2.3205 | 150 | T3(Correct) | |

Fundamental ferroresonance | 0.5990 | 1.7498 | 0.5775 | 3.1256 | 0 | T4(Correct) |

0.6413 | 1.7021 | 0.5954 | 1.9397 | 0 | T4(Correct) | |

0.5265 | 1.6828 | 0.7097 | 7.0827 | 150 | T2(Incorrect) | |

Sub-frequency ferroresonance | 0.4020 | 1.0328 | 1.5833 | 2.1671 | 25 | T5(Correct) |

0.4223 | 1.0977 | 1.5727 | 1.9389 | 25 | T5(Correct) | |

0.4233 | 1.1202 | 1.5081 | 2.3553 | 25 | T5(Correct) | |

Single phase-to-ground | 0.6200 | 1.7281 | 0.0126 | 2.4050 | 0 | T6(Correct) |

0.6173 | 1.7276 | 0.2456 | 2.3049 | 0 | T6(Correct) | |

0.6181 | 1.7284 | 0.0126 | 2.3999 | 0 | T6(Correct) | |

Line switching | 0.5997 | 1.0096 | 0.9854 | 13.8765 | 10 | T7(Correct) |

0.5845 | 1.0046 | 0.9786 | 20.8586 | 0 | T7(Correct) | |

0.6604 | 1.0152 | 1.0095 | 27.7779 | 0 | T7(Correct) | |

Lightning | 0.5533 | 1.0160 | 1.0235 | 17.5915 | 0 | T7(Correct) |

0.6003 | 0.9902 | 0.9920 | 13.4594 | 0 | T7(Correct) | |

0.5756 | 0.9759 | 0.9792 | 18.1713 | 0 | T7(Correct) |

Type | Rate (%) |
---|---|

Capacitor switching | 95.0 |

Arc grounding | 97.1 |

High frequency ferroresonance | 100.0 |

Fundamental ferroresonance | 97.0 |

Sub-frequency ferroresonance | 100.0 |

Single phase-to-ground | 100.0 |

Line switching | 95.2 |

Lightning | 97.9 |

Overall | 97.8 |

## 6. Conclusions

## Acknowledgements

## References

- Juan, L.; Chenching, L.; Schneider, K.P. Controlled partitioning of a power network considering real and reactive power balance. IEEE Trans. Smart Grid
**2010**, 1, 261–269. [Google Scholar] [CrossRef] - Pei, Z.; Fangxing, L.; Bhatt, N. Next-generation monitoring, analysis, and control for the future smart control center. IEEE Trans. Smart Grid
**2010**, 1, 186–192. [Google Scholar] [CrossRef] - Mokryani, G.; Siano, P.; Piccolo, A. Identification of ferroresonance based on S-transform and support vector machine. Simul. Model. Pract. Theory
**2010**, 18, 1412–1424. [Google Scholar] [CrossRef] - Mokryani, G.; Haghifam, M.R.; Esmaeilpoor, J. Identification of ferroresonance based on wavelet transform and artificial neural network. Eur. Trans. Electr. Power
**2009**, 19, 474–486. [Google Scholar] [CrossRef] - Gu, Y.H.; Bollen, M.H.J. Time-frequency and time-scale domain analysis of voltage disturbances. IEEE Trans. Power Deliv.
**2000**, 15, 1279–1284. [Google Scholar] [CrossRef] - Gaing, Z.L. Wavelet-based neural network for power disturbance recognition and classification. IEEE Trans. Power Deliv.
**2004**, 19, 1560–1568. [Google Scholar] [CrossRef] - Kezunovic, M.; Yuan, L. A novel software implementation concept for power quality study. IEEE Trans. Power Deliv.
**2002**, 17, 544–549. [Google Scholar] [CrossRef] - Stockwell, R.G.; Mansinha, L.; Lowe, R.P. Localization of the complex spectrum: the S Transform. IEEE Trans. Signal Process
**1996**, 44, 998–1001. [Google Scholar] [CrossRef] - Fengzhan, Z.; Rengang, Y. Power-quality disturbance recognition using s-transform. IEEE Trans. Power Deliv.
**2007**, 22, 944–950. [Google Scholar] - Dash, P.K.; Panigrahi, B.K.; Panda, G. Power quality analysis using s-transform. IEEE Trans. Power Deliv.
**2003**, 18, 406–411. [Google Scholar] [CrossRef] - Mishra, S.; Bhende, C.N.; Panigrahi, B.K. Detection and classification of power quality disturbances using s-transform and probabilistic neural network. IEEE Trans. Power Deliv.
**2008**, 23, 280–287. [Google Scholar] [CrossRef] - Àkànbí, L.A.; Odéjobí, O.À. Automatic recognition of oral vowels in tone language: Experiments with fuzzy logic and neural network models. Appl. Soft Comput.
**2011**, 11, 1467–1480. [Google Scholar] - Liao, Y.; Lee, J.B. A fuzzy-expert system for classifying power quality disturbances. Int. J. Electr. Power
**2004**, 26, 199–205. [Google Scholar] - Monsef, H.; Ranjbar, A.M.; Jadid, S. Fuzzy rule-based expert system for power system fault diagnosis. IEE Proc. Generat. Transm. Distrib.
**1997**, 144, 186–192. [Google Scholar] - Yen, V.C. Rule selections in fuzzy expert systems. Expert Syst. Appl.
**1999**, 16, 79–84. [Google Scholar] [CrossRef] - Zadeh, L.A. Fuzzy sets. Inf. Control
**1965**, 8, 338–353. [Google Scholar] [CrossRef] - Cortes, C.; Vapnik, V. Support-vector networks. Mach. Learn.
**1995**, 20, 273–297. [Google Scholar] - Donoho, D.L. De-noising by soft-thresholding. IEEE Trans. Inf. Theory
**1995**, 41, 613–627. [Google Scholar] - Hao, Z.; Blackburn, T.R.; Phung, B.T.; Sen, D. A novel wavelet transform technique for on-line partial discharge measurements. Part 1: WT de-noising algorithm. IEEE Trans. Dielectr. Electr. Insul
**2007**, 14, 3–14. [Google Scholar]

© 2011 by the authors; licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution license (http://creativecommons.org/licenses/by/3.0/).

## Share and Cite

**MDPI and ACS Style**

Wang, J.; Yang, Q.; Sima, W.; Yuan, T.; Zahn, M. A Smart Online Over-Voltage Monitoring and Identification System. *Energies* **2011**, *4*, 599-615.
https://doi.org/10.3390/en4040599

**AMA Style**

Wang J, Yang Q, Sima W, Yuan T, Zahn M. A Smart Online Over-Voltage Monitoring and Identification System. *Energies*. 2011; 4(4):599-615.
https://doi.org/10.3390/en4040599

**Chicago/Turabian Style**

Wang, Jing, Qing Yang, Wenxia Sima, Tao Yuan, and Markus Zahn. 2011. "A Smart Online Over-Voltage Monitoring and Identification System" *Energies* 4, no. 4: 599-615.
https://doi.org/10.3390/en4040599