Series arc fault on-line detection method and device

Abstract

The application discloses a series fault arc online detection method and device, the method is to use fuzzy control algorithm to fuse the series fault arc comprehensive feature recognition coefficient obtained by using the power circuit current zero rest time coefficient and the normalized variance coefficient, so as to judge whether the series fault arc occurs or not; the device is composed of a load current collection module, a voltage signal collection module, a current signal conditioning module, a voltage signal conditioning module, a single-chip microcomputer module, a DSP unit, an optoelectronic coupler module, a reset circuit module, an Ethernet communication unit, an alarm unit and a control output unit. The application can accurately judge whether the series fault arc occurs or not, and does not have the defects of false detection, false protection, missed detection or missed protection, and the algorithm is simple and easy to implement in engineering.

Description

TECHNICAL FIELD

[0001] The present application belongs to the field of electric safety, and relates to an online detection method and device for low-voltage series arc fault. BACKGROUND

[0002] In an effective grounding single-phase alternating current power distribution system, the insulation performance of the cable and load insulation layer in the power supply circuit is reduced, and the electrical looseness causes series arc fault and even electrical fire hazard, which threatens the safety of people's life and property.

[0003] At present, the detection methods for series arc fault known in the art mainly include three types. The first type is a method for detecting corresponding electric parameters based on an arc mathematical model to identify the arc. This method is slow in application and development due to the imperfection of the established arc mathematical model, the large number of detection parameters, and the complexity of the arc identification algorithm. The second type is a method for indirectly identifying the arc based on the physical phenomena such as sound, light, heat and electromagnetic radiation caused by the arc discharge. This method is limited in application due to the uncertain arc occurrence position and the diversified types of electric loads, and has certain limitations in application scope. The third type is a method for identifying the arc based on the characteristic changes of the circuit voltage and current signals when the arc fault occurs. The voltage signal characteristic change identification method is difficult to effectively sample the voltage due to the large range of load position changes, and is hindered in practical application. The current signal characteristic change identification method includes a method based on one or more characteristics of the series arc current for fault identification (such as time domain and frequency domain characteristics, wavelet transform analysis method, kurtosis and pulse number double criterion method), a method based on the overall characteristics of the series arc fault current for fault identification (such as neural network analysis method), and a method based on the fusion technology of one or more characteristics and overall characteristics of the series arc current for fault identification (such as a method based on support vector machine technology after classification and sorting of arc current characteristic quantities, a method based on correlation coefficient and skewness index for identifying arc fault). These methods have defects such as large algorithm difficulty, poor real-time performance, and low fault identification accuracy. SUMMARY

[0004] The present application aims to overcome the above-mentioned defects of the existing methods, and provides an online detection method for series arc fault, and also provides an online detection protection device for series arc fault which can be used to realize the method.

[0005] A series arc fault detection protection device, characterized in that it comprises a load current signal acquisition module, a current signal conditioning module, a voltage signal acquisition module, a voltage signal conditioning module, a control output unit, a single-chip microcomputer module and a DSP unit; further comprising an Ethernet communication unit, a reset circuit module and an alarm unit connected with the single-chip microcomputer module, and the single-chip microcomputer module is connected with an upper computer through the Ethernet communication unit.

[0006] Specifically, the output terminal of the load current signal acquisition module is connected to the input terminal of the current signal conditioning module; the output terminal of the voltage signal acquisition module is connected to the input terminal of the voltage signal conditioning module; the output terminals of the current signal conditioning module and the voltage signal conditioning module are connected to the input terminal of the DSP unit; the output terminal of the DSP unit is connected to the input terminal of the microcontroller module; and the output terminal of the microcontroller module is connected to the control output unit.

[0007] The microcontroller module is connected to the alarm unit and the control output unit through the optocoupler module. It is used to identify the series fault arc signal and issue a power-off control command. The control output unit is used to cut off the power supply circuit of the load after receiving the power-off control command.

[0008] The load current signal acquisition module is a current transformer.

[0009] The current signal conditioning module consists of an I / V converter, an RC filter circuit, an amplifier circuit, and a limiting circuit connected in sequence.

[0010] The voltage signal acquisition module is a voltage transformer.

[0011] The voltage signal conditioning module consists of an I / V converter, an RC filter circuit, an amplifier circuit, and a limiting circuit connected in sequence.

[0012] A method for online detection of series fault arcs includes the following implementation steps:

[0013] 1. The instantaneous values ​​x of two cycles of current signals in the power circuit are continuously acquired through the load current signal acquisition module. i y i The effective value Ix is calculated by the DSP unit according to formulas (1) and (2) respectively. rms 、Iy rms Calculate the zero-rest time t according to formulas (3) and (4). x t y Calculate the normalized zero-rest time coefficient A1 according to formula (5).

[0014]

[0015]

[0016]

[0017]

[0018]

[0019] In the formula x i y iThese are the i-th sampled values ​​of the current waveforms x and y in the power supply circuit, respectively; k is the total number of sampling points in one cycle of the current waveform, and f s The sampling frequency of the current waveform;

[0020] 2. The DSP unit calculates x in the power circuit according to formula (6). i y i The difference R i Then, calculate its root mean square error G using formula (7), and calculate the normalized variance coefficient A2 of G using formula (8).

[0021] R i =x i -y i i = 1, 2, ..., N (6)

[0022]

[0023]

[0024] In the formula, A2 is the normalized variance coefficient;

[0025] Third, the DSP unit performs fuzzy fusion of the two feature variables A1 and A2 according to the fuzzy control algorithm to obtain the comprehensive feature identification coefficient A0 of the series fault arc.

[0026] Specifically, it also includes performing fuzzy reasoning on A1, A2, and A0 according to formula (9) to obtain the membership degree u of element z belonging to A0. A0′ (z), according to formula (10) for u A0′ (z) After obtaining A0 through defuzzification, it is further compared with an empirical threshold l (usually taken as 0.5). If A0 is less than l, it is determined that a series fault arc has occurred; otherwise, no series fault arc has occurred.

[0027] u A0′ (z)={∨ m∈V u A1′(m) ∧(u A1 (m)∧u A 0(z))}∧{∨ n∈V u A2′(n) ∧(u A (2n)∧u A (0z))}(9)

[0028]

[0029] In the formula, m is an element in the A1 domain, u A1 (m) represents the membership degree of m to A1, u A1′ (m) represents the membership degree of the new input element to A1, n represents the element in the universe of discourse A2, and uA2 (n) is the membership of n belonging to A2, u A2′ (n) is the membership of input new element belonging to A2', z is an element in A0 universe, u A0 (z) is the membership of z belonging to A0, u A0′ (z) is u A1′ (m) and u A2′ (n) fuzzy identification result.

[0030] The present application has the following beneficial effects:

[0031] I. The method and device adopt fuzzy fusion to obtain a series fault arc comprehensive feature identification coefficient A0 by using a current zero rest time coefficient A1 and a normalized variance coefficient A2 in an electric circuit, so that the occurrence of the series fault arc can be accurately judged, and there is no defect of false detection or false protection or missed detection or missed protection, and the algorithm is simple and easy to implement in engineering;

[0032] II. The method and device can also simultaneously perform online detection on load overcurrent, overvoltage and under-voltage, online monitoring and statistics on load energy consumption, and remote control on load opening and closing, so that one machine has multiple uses and the cost performance is high. BRIEF DESCRIPTION OF DRAWINGS

[0033] Figure 1 Fig. 1 is a structural schematic diagram of the series fault arc online detection device;

[0034] Figure 2 Fig. 2 is a membership function diagram of the input zero rest time coefficient A1 on fuzzy variables;

[0035] Figure 3 Fig. 3 is a membership function diagram of the input normalized variance coefficient A2 on fuzzy variables;

[0036] Figure 4 Fig. 4 is a membership function diagram of the output series fault arc comprehensive feature identification coefficient A0 on fuzzy variables. DETAILED DESCRIPTION

[0037] The structural implementation of the series fault arc online detection protection device is shown in the accompanying drawings. Figure 1As shown, it includes load current signal acquisition module, current signal conditioning module, voltage signal acquisition module, voltage signal conditioning module, reset circuit module, single-chip microcomputer module, optoelectronic coupler module, DSP unit, Ethernet communication unit, alarm unit, control output unit. Among them, the output end of the load current signal acquisition module is connected with the input end of the current signal conditioning module; the output end of the voltage signal acquisition module is connected with the input end of the voltage signal conditioning module; the output end of the current signal conditioning module and the voltage signal conditioning module is connected with the input end of the DSP unit, the output end of the DSP unit is connected with the input end of the single-chip microcomputer module, and the output end of the single-chip microcomputer module is connected with the control output unit.

[0038] The load current signal acquisition module is a current transformer, and the current signal conditioning module is composed of I / V conversion, RC filter circuit, amplification circuit and limiting circuit connected in sequence.

[0039] The voltage signal acquisition module is a voltage transformer, and the voltage signal conditioning module is composed of I / V conversion, RC filter circuit, amplification circuit and limiting circuit connected in sequence.

[0040] The application also includes an Ethernet communication unit, a reset circuit module and an alarm unit connected with the single-chip microcomputer module, and the single-chip microcomputer module is connected with the upper computer through the Ethernet communication unit, wherein the single-chip microcomputer module is connected with the alarm unit and the control output unit through the optoelectronic coupler module, for identifying the series fault arc signal and issuing a power-off control command, and the control output unit cuts off the power supply of the load supply loop after receiving the power-off control command. The Ethernet communication unit uploads the fault information in the loop to the upper computer.

[0041] In the structural design of the specific example, the load current signal detection module adopts a current transformer (such as HCT255A, HRCT-1), the voltage signal detection module adopts a voltage transformer (such as HPT304A, HRPT-1), the single-chip microcomputer system module can adopt a digital signal processor (such as DSPIC30F6014A) with built-in multi-channel A / D, the DSP unit can adopt a TSM320 series device (such as TSM320F2812), the current signal conditioning module and the voltage signal conditioning module are realized by a filter composed of I / V conversion and amplification circuit (such as RVC420, LM324 and peripheral auxiliary elements), clamping diode (such as IN4148), limiting voltage stabilizing tube and general resistance-capacitance device. The optoelectronic coupler module needs to select an optocoupler device (such as TLP127) with large output driving capacity. The control output module can adopt a general control relay (such as JQX-14) or a solid-state relay SSR (such as S310ZK), and the alarm unit adopts a buzzer or an indicator light.

[0042] In addition, when the series fault arc is excluded, the single-chip microcomputer module is reset by the reset circuit module.

[0043] The series fault arc online detection method comprises the following implementation steps:

[0044] I. Continuously collecting two-cycle current signal instantaneous values x i , y i of the power circuit by the load current signal acquisition module rms , Iy rms , calculating zero-hibernation time t x , t y according to formulas (3) and (4), and calculating the normalized zero-hibernation time coefficient A1

[0045]

[0046]

[0047]

[0048]

[0049]

[0050] wherein x i , y i are the i-th sampling values of the current waveforms x and y of the power circuit, k is the total sampling point number of one cycle of the current waveform, f s is the current waveform sampling frequency;

[0051] IV. Calculating the difference R i of x i and y i in the power circuit by the DSP unit according to formula (6), calculating the mean square deviation G of R

[0052] R i = x i -y i i=1,2…,N (6)

[0053]

[0054]

[0055] wherein A2 is the normalized variance coefficient;

[0056] 5. The DSP unit performs fuzzy fusion of the two feature variables A1 and A2 according to the fuzzy control algorithm to obtain the comprehensive feature identification coefficient A0 of the series fault arc.

[0057] Specifically, fuzzy sets are defined according to Table 1 below.

[0058] Table 1 Fuzzy Set of Variables

[0059]

[0060] Specifically, the membership degree of each element in the domain to the fuzzy variable is determined according to... Figure 2 , Figure 3 , Figure 4 Confirmed (this is a custom design based on extensive experiments and experience).

[0061] Specifically, the fuzzy recognition rule table is defined as shown in Table 2 (it is a custom definition based on a large number of experiments and experience).

[0062] Table 2 Fuzzy Recognition Rules

[0063]

[0064] Specifically, it also includes performing fuzzy reasoning on A1, A2, and A0 according to formula (9) to obtain the membership degree u of element z belonging to A0. A0′ (z), according to formula (10) for u A0′ (z) After obtaining A0 through defuzzification, it is further compared with an empirical threshold l (usually taken as 0.5). If A0 is less than l, it is determined that a series fault arc has occurred; otherwise, no series fault arc has occurred.

[0065] u A0′ (z)={∨ m∈V u A1′(m) ∧(u A1 (m)∧u A0 (z))}∧{∨ n∈V u A2′(n) ∧(u A2 (n)∧u A0 (z))} (9)

[0067]

[0068] In the formula, m is an element in the A1 domain, u A1 (m) represents the membership degree of m to A1, u A1′ (m) represents the membership degree of the new input element to A1, n represents the element in the universe of discourse A2, and u A2 (n) represents the membership degree of n to A2, u A2′(n) is the membership of the input new element belonging to A2', z is an element in A0 universe, u A0 (z) is the membership of z belonging to A0, u A0′ (z) is u A1′ (m) and u A2′ (n) fuzzy recognition result.

[0069] The basic content of the series fault arc online detection method of the embodiment is that the voltage signal acquisition module continuously and synchronously acquires the power circuit voltage u, the load current signal acquisition module acquires the load current signal instantaneous value i, after being amplified, limited, filtered and I / V converted by the current signal conditioning module and the voltage signal conditioning module, the load current signal instantaneous value i is sent to the DSP unit containing a built-in A / D converter, and the current zero rest time of the power circuit is calculated and stored by the DSP unit according to the corresponding calculation formula.

[0070] The difference value and the mean square error of the two period currents are calculated by the DSP unit according to formulas (6) and (7) respectively.

[0071] The current zero rest time of the power circuit and the mean square error obtained from formula (7) are normalized by the DSP unit according to formulas (5) and (8), and two characteristic variables A1 and A2 are calculated.

[0072] u (z) is calculated by the DSP unit according to formula (9), and the series fault arc comprehensive feature recognition coefficient A0 is calculated according to formula (10) and sent to the single-chip microcomputer module, and the single-chip microcomputer module takes A0 as the basis for series fault arc detection. A0′ (z), and the series fault arc comprehensive feature recognition coefficient A0 is calculated according to formula (10) and sent to the single-chip microcomputer module, and the single-chip microcomputer module takes A0 as the basis for series fault arc detection.

[0073] The structure implementation of the series fault arc online detection device is shown in the accompanying drawings. Figure 2 The structure implementation of the series fault arc online detection device is shown in the accompanying drawings.

[0074] The load current signal acquisition module is a current transformer, and the current signal conditioning module is composed of I / V conversion, RC filter circuit, amplification circuit and limiting circuit connected in sequence.

[0075] The voltage signal acquisition module is a voltage transformer, and the voltage signal conditioning module is composed of I / V conversion, RC filter circuit, amplification circuit and limiting circuit connected in sequence.

[0076] The application also comprises an Ethernet communication unit, a reset circuit module and an alarm unit connected with the single-chip microcomputer module, and the single-chip microcomputer module is connected with the upper computer through the Ethernet communication unit, wherein the single-chip microcomputer module is connected with the alarm unit and the control output unit through the optoelectronic coupler module, used for identifying the series fault arc signal and issuing a power-off control command, and the control output unit cuts off the power supply of the load power supply loop after receiving the power-off control command. The Ethernet communication unit uploads the fault information in the loop to the upper computer.

[0077] In the structural design of the specific example, the load current signal detection module adopts a current transformer (such as HCT255A, HRCT-1), the voltage signal detection module adopts a voltage transformer (such as HPT304A, HRPT-1), the single-chip microcomputer system module can adopt a digital signal processor (such as DSPIC30F6014A) with built-in multi-channel A / D, the DSP unit can adopt a TSM320 series device (such as TSM320F2812), the current signal conditioning module and the voltage signal conditioning module are realized by a filter composed of I / V conversion and amplification circuit (such as RVC420, LM324 and peripheral auxiliary elements), clamping diode (such as IN4148), limiting voltage stabilizing tube and general resistance-capacitance device. The optoelectronic coupler module needs to select an optocoupler device (such as TLP127) with large output driving capacity. The control output module can adopt a general control relay (such as JQX-14) or a solid-state relay SSR (such as S310ZK), and the alarm unit adopts a buzzer or an indicator lamp.

[0078] In addition, when the series fault arc is excluded, the single-chip microcomputer module is reset through the reset circuit module.

[0079] The working process of the device for detecting the series fault arc online is as follows:

[0080] I. System power-on, self-checking and initialization;

[0081] II. Setting and modifying of corresponding detection parameters according to actual operation requirements;

[0082] III. Checking whether the device itself has a fault, and if yes, entering a corresponding processing program and giving a fault diagnosis prompt information, and if not, executing the next step;

[0083] Four, continuously collect the voltage across the load to be tested, the load current value;

[0084] Five, real-time calculation and storage of x i , y i , t x , t y , R i , and fuse A1 and A2 into A0;

[0085] Six, judge whether A0 is greater than l (generally 0.5), if A0 is less than l, it is determined that a series fault arc is detected, and step seven is directly executed, if A0 is greater than or equal to l, it is considered that a series fault arc does not occur, and step eight is executed;

[0086] Seven, control the output unit to cut off the load power supply circuit, and the buzzer of the alarm unit alarms;

[0087] Eight, the Ethernet communication unit uploads the series fault arc information in the load circuit to the upper computer for remote monitoring;

[0088] Nine, repeat step three.

[0089] The basic principles, main features and advantages of the present application are described through the above description and examples, in particular, the essence of the present application is shown, that is, the series fault arc comprehensive feature recognition coefficient is obtained by using the fuzzy fusion algorithm to fuse the current zero rest time and the normalized variance coefficient, when the series fault arc comprehensive feature recognition coefficient is less than the empirical threshold, it is judged that a series fault arc occurs, so as to realize accurate and effective online detection of the series fault arc. The practical application mode and application occasion of the present application are not limited by the above examples, various corresponding changes (such as the application occasion changes to the effective grounding power supply system and its power supply cabinet (box)) and improvements of the present application can be made without departing from the scope of the present application, these changes and improvements should fall within the scope of the present application, and the protection scope can be defined by the claims attached to the present application and their equivalents.

Claims

1. A series arc fault online detection method, the device used comprises a load current signal acquisition module, a current signal conditioning module, a voltage signal acquisition module, a voltage signal conditioning module, a control output unit, a single-chip microcomputer module and a DSP unit; further comprising an Ethernet communication unit, a reset circuit module and an alarm unit connected with the single-chip microcomputer module, and the single-chip microcomputer module is connected with a host computer through the Ethernet communication unit; characterized in that, The implementation steps specifically include the following: Step one, continuously collect two cycle current signal instantaneous value in the power circuit through the load current signal acquisition module x i , y i , calculate its effective value according to formula 1 and formula 2 by DSP unit respectively Ix rms , Iy rms , calculate zero rest time according to formula 3 and formula 4 t x , t y , calculate the normalized zero rest time coefficient according to formula 5 A1 , Equation 1 Equation 2 Equation 3 Equation 4 , Equation 5 wherein x i , y i is the current waveform of the power circuit x, y the first sample value of the current waveform i ; k is the total number of sample points of a period of the current waveform, f s is the sampling frequency of the current waveform; Step two, the DSP unit calculates the difference between the two values of the power circuit according to formula 6 x i , y i R i G G A2 ,​​​​ Formula 6 Equation 7 Formula 8 In the formula, A2 is the normalized coefficient of variance; Step three, the DSP unit fuses the two characteristic variables according to the fuzzy control algorithm to obtain a comprehensive characteristic identification coefficient of series fault arc A1 , A2 A0 ,​ In particular, it also includes the determination of A1 , A2 , A0 The fuzzy inference is performed according to formula 9 to obtain The membership degree of the element to A0 The inverse fuzzy processing is performed on to obtain A0 After that, it is further compared with the experience threshold value. If A0 is less than , it is determined that the series fault arc occurs, otherwise, the series fault arc does not occur.​ Equation 9 Formula 10 wherein m is A1 an element of the domain of discourse, is m a membership of A1 to a membership of the new element input to A1 a membership of A2 an element of the domain of discourse, is a membership of A2 to a membership of the new element input to A a membership of A0 an element of the domain of discourse, is a membership of A0 to is and fuzzy recognition result.

Images