JP2013099230A - Independent operation detection device for system interaction inverter device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To detect an independent operation state of a system interaction inverter device at high speed with high precision.SOLUTION: An independent operation detection device 7 inputs a variation control signal varying in amplitude in proportion to a variation rate of a frequency generated by a variation control signal generator 71 to a controller 6 to vary the amplitude of the output voltage of an inverter 2 so that the variation amount varied according to the variation rate of the frequency. The independent operation detection device 7 removes an unbalanced component and harmonic components included in a three-phase output voltage of the inverter 2 detected by an AC voltage detector 8 by a complex coefficient band-pass filter, and detects the frequency using only a fundamental wave component output from the complex coefficient band-pass filter. Then a variation amount of the frequency is found, and when the variation amount exceeds a predetermined threshold, the inverter 2 is considered to enter an independent operation state, so that a breaker 5 disconnects a system interaction inverter device A from an electric power system B.

Description

  The present invention relates to an isolated operation detection device that detects an isolated operation of a grid-connected inverter device linked to an electric power system.

  When the distributed power source connected to the power system is connected to the high-voltage distribution system under the condition of reverse power flow, an isolated operation detection device is provided to prevent the isolated operation. As a method for detecting the single operation state of the distributed power source, frequency fluctuations that occur during single operation such as frequency change rate detection method, QC mode frequency shift method, slip mode frequency shift method, active power fluctuation method, reactive power fluctuation method, etc. Various systems are known for detecting islanding by detecting. The frequency change rate detection method is a passive method, but the slip mode frequency shift method, the QC mode frequency shift method, the active power fluctuation method, and the reactive power fluctuation method are active methods.

  The frequency change rate detection method is a method for detecting a sudden change in the output frequency of the distributed power source due to an imbalance between the output of the distributed power source and the load. In the slip mode frequency shift method, when the C (capacitance) component of the reactive power load is large during single operation, the output frequency of the distributed power source increases, and when the L (inductance) component is large, the output frequency of the distributed power source decreases. Using this characteristic, when the output frequency of the distributed power source fluctuates from the reference frequency, the amount of fluctuation of the output frequency is amplified and detected. Specifically, the PCS (Power Conditioning System) has frequency-phase characteristics, and when the output frequency of the distributed power source rises with respect to the reference frequency, the phase of the output current of the PCS is advanced and output by positive feedback. When the frequency increase is accelerated and the output frequency of the distributed power source decreases with respect to the reference frequency, the phase of the output current of the PCS is delayed and the decrease in the output frequency is accelerated by positive feedback. This is a method of detecting islanding by the amount of fluctuation.

  The QC mode frequency shift method uses the above frequency characteristics as in the slip mode frequency shift method, but amplifies the phase of the output current of the PCS by positively feeding back the rate of change of the output frequency instead of the output frequency. It is a method to make it.

  The active power variation method is a method in which periodic active power variation is given to the output of the PCS, and the presence or absence of independent operation is detected based on the variation amount of the output voltage, output current, and output frequency of the PCS. In addition, the reactive power fluctuation method is a method in which periodic reactive power fluctuations are given to the output of the PCS, and the presence or absence of an isolated operation is detected based on the fluctuation amounts of the PCS output voltage, output current, and output frequency.

JP 11-41820 A JP 2000-358331 A Japanese Patent Laid-Open No. 2002-281684 Japanese Patent Laid-Open No. 2007-252127 Japanese Patent Laid-Open No. 2011-30306 JP 2006-25550 A

"PLL and frequency detection method to cope with abnormal voltage in case of power system failure" Denki Theory B, Vol. 118, No. 9, 1998

  In a distributed power source using a grid-connected inverter device, when an isolated operation state is detected, for example, it is desired to disconnect the grid-connected inverter device from a customer (load) within 1 second or to stop the operation. ing. The method using the output frequency of the distributed power source as a detection parameter in the above isolated operation detection method requires a high-speed and high-accuracy frequency detection device.

  On the other hand, as a method to detect the output frequency of the distributed power supply, the instantaneous value of the AC voltage output from the distributed power supply is detected, and the time between the points where the detected value crosses the zero level (zero cross point) is measured. Thus, a method of detecting the output frequency of the distributed power supply (zero cross point counting method) is known (see Patent Document 6). Non-Patent Document 1 proposes a method for obtaining the frequency of the AC voltage using a phase detection device that detects the phase of the AC voltage output from the distributed power source using a multiplying PLL (Phase Locked Loop). Has been.

  In the conventional frequency detection device using the zero-cross point counting method, the frequency calculation processing can be performed only at the timing when the sampling value of the detection voltage detects the zero-cross point. For example, the frequency is detected every time the detection voltage is sampled. Cannot detect the continuous frequency. For this reason, if the phase of the power system changes suddenly during a period when frequency detection is not possible, or if a voltage shortage occurs due to a ground fault, the frequency cannot be detected accurately (immediate response and detection accuracy are not good). There's a problem.

  In addition to the fundamental positive phase component (hereinafter simply referred to as “fundamental wave component”), the frequency of the three-phase power system (hereinafter referred to as “system frequency”) includes low-order harmonic components (for example, 5th, 7th and 11th order harmonic components (hereinafter simply referred to as “harmonic components”) and fundamental wave antiphase components (hereinafter referred to as “unbalanced components”) are often included. If the detected voltage is included in the detected voltage, the waveform of the detected voltage does not accurately become the waveform of the fundamental wave component (waveform distortion occurs), and the zero-cross point of the fundamental wave component cannot be accurately detected. There is also a problem that the detection accuracy of the is reduced.

  On the other hand, the method of detecting the frequency using the multiplying PLL phase detector does not directly detect the frequency of the AC signal output from the distributed power supply, and thus has a problem in terms of detection accuracy. A phase detection device using a PLL generates a phase in the device, calculates a phase difference between the phase and the phase of the input AC signal, and controls the generated phase so that the phase difference becomes zero. Therefore, when a phase difference occurs, the phase generated in the phase detection device deviates from the detection target phase. Therefore, in the method of calculating the frequency from the phase generated in the phase detection device, there is a problem in terms of detection accuracy and response speed when the system frequency varies.

  Therefore, when the frequency detection device using the zero cross point counting method is applied to the isolated operation detection device, there are problems in terms of detection continuity and detection accuracy. Also. There is a problem in terms of detection accuracy and response speed even when a method of detecting a frequency using a multiplying PLL phase detection device is applied to an isolated operation detection device.

  The present invention has been conceived under the circumstances described above, and an object thereof is to provide an isolated operation detection device capable of detecting the isolated operation state of a grid-connected inverter device at high speed and with high accuracy. And

  The isolated operation detection device for a grid-connected inverter device according to claim 1 is a frequency detection means for detecting a frequency of an AC signal output from at least a grid-connected inverter device linked to a power system, and the frequency detection Independent operation detection means for detecting that the grid-connected inverter device has shifted to the single operation state based on the detection value of the means, the frequency detection means, AC signal detecting means for detecting the AC signal, and a complex coefficient filter for removing unbalanced components and harmonic components contained in the AC signal detected by the AC signal detecting means and outputting only the fundamental wave component. Using the first filter means and the fundamental wave component of the AC signal output from the first filter means, the frequency of the fundamental wave component is calculated. Characterized in that it comprises a frequency calculating means.

  The isolated operation detection apparatus for a grid-connected inverter device according to claim 1, wherein the complex coefficient filter of the first filter means is a complex coefficient bandpass filter whose center frequency is set to the frequency of the AC signal. Good (claim 2).

  Further, in the isolated operation detection apparatus for a grid-connected inverter device according to claim 1 or 2, the frequency calculation means sets the center frequency to the frequency of the AC signal, and the phase difference linearly changes in the pass band. Second filter means comprising a multi-coefficient bandpass filter having phase characteristics, and a fundamental wave component of the AC signal input to the second filter means and the AC signal output from the second filter means Phase difference calculating means for calculating a phase difference of the fundamental wave component of the AC signal generated in the second filter means by a predetermined calculation formula using a fundamental wave component of the phase difference, and a phase difference in the passband based on the phase characteristic The frequency of the fundamental wave component of the AC signal input to the second filter means is calculated using a predetermined relational expression for obtaining the frequency from the phase difference and the phase difference. Better to configuration including a frequency computing unit, the (claim 3).

  The frequency calculation means includes a second filter means comprising a multi-coefficient bandpass filter having a phase characteristic in which a center frequency is set to a frequency of the AC signal and a phase difference linearly changes in a pass band; Generated by the second filter means by a predetermined arithmetic expression using the fundamental wave component of the AC signal input to the second filter means and the fundamental wave component of the AC signal output from the second filter means. A phase difference calculating means for calculating a phase difference of a fundamental wave component of the AC signal, a predetermined table indicating a relationship between a phase difference and a frequency in the pass band set in advance based on the phase characteristic, and the table from the table By reading the frequency corresponding to the phase difference calculated by the phase difference calculating means, the frequency of the fundamental wave component of the AC signal input to the second filter means is read. A frequency calculating means for calculating the number, may be a configuration including (claim 4).

  The frequency calculation means has a phase characteristic in which the phase difference is zero at the center frequency, becomes negative in a frequency region larger than the center frequency, and becomes positive in a small frequency region, and the center frequency can be changed. Second filter means composed of a multi-coefficient filter having a large passband, a fundamental wave component of the AC signal input to the second filter means, and a basic of the AC signal output from the second filter means A phase difference calculation means for calculating a phase difference of the fundamental wave component of the AC signal generated by the second filter means by a predetermined calculation formula using a wave component, and a phase difference calculated by the phase difference calculation means is positive. If the phase difference calculated by the phase difference calculating means is negative, the center frequency of the second filter means is decreased by a predetermined change amount until the phase difference becomes zero. Center frequency control means for raising the center frequency of the second filter means by the amount of change until the value becomes zero and outputting the changed center frequency as a detected value of the frequency. 5).

  6. The islanding operation detection device for a grid-connected inverter device according to claim 5, wherein the complex coefficient filter of the first filter means is based on a center frequency output from the center frequency control means with respect to the center frequency. It is preferable that the multi-coefficient notch filter has a stop frequency controlled so as to be a negative frequency component and a harmonic component of a predetermined order.

  The isolated operation detection device for a grid-connected inverter device according to any one of claims 1 to 6, wherein the grid-connected inverter device is a power that controls a reactive power amount output from the grid-connected inverter device. Reactive power having a major loop, generating a reactive power fluctuation value that fluctuates the reactive power amount based on the frequency of the AC signal detected by the frequency detection means, and feeding back to the power major loop The system further includes a fluctuation value generating means, wherein the islanding operation detecting means calculates a fluctuation amount of the frequency detected by the frequency detecting means, and the grid interconnection inverter device alone is calculated when the fluctuation amount exceeds a predetermined threshold value. It may be detected that the operation state has been shifted (claim 7).

  8. The isolated operation detection apparatus for a grid-connected inverter device according to claim 7, wherein the reactive power fluctuation value generation means calculates a frequency change rate using the frequency of the AC signal detected by the frequency detection means. Then, the reactive power fluctuation value that varies in proportion to the frequency change rate may be generated.

  Further, in the isolated operation detection apparatus for a grid-connected inverter device according to any one of claims 1 to 8, the AC signal is a single-phase or three-phase AC signal (Claim 9).

  According to the present invention, the unbalanced component and the harmonic component are removed by passing the detected value of the AC signal output from the grid interconnection inverter device through the complex coefficient filter, and only the fundamental component is extracted. Since the frequency of the wave component is calculated, the frequency of the AC signal can be detected at high speed and with high accuracy. Therefore, the isolated operation of the grid-connected inverter device can be detected with high speed and high accuracy with a simple configuration.

  In particular, when a complex coefficient notch filter is used as a complex coefficient filter, unbalanced components and higher harmonic components can be removed more favorably than when a bandpass filter is used as a complex coefficient filter, and AC with higher speed and higher accuracy. The frequency of the signal can be detected.

  Moreover, if it is a system which detects the frequency of the alternating current signal output from a grid connection inverter apparatus, and detects an isolated operation based on the fluctuation | variation of the frequency, regardless of a passive system and an active system, A single operation of the grid-connected inverter device can be detected at high speed with high accuracy. Moreover, independent operation of the grid-connected inverter device can be detected at high speed with high accuracy regardless of whether the power system is a single-phase power system or a three-phase power system.

It is a figure which shows the structure of the grid connection inverter apparatus provided with the isolated operation detection apparatus which concerns on this invention. It is a block diagram which shows the basic composition of the process for producing | generating the PWM signal in a control apparatus. It is a figure which shows the block configuration of 1st Embodiment of a frequency detection part. It is a figure for demonstrating the relationship between the fundamental wave component of a three-phase alternating voltage, and an unbalanced component. It is a figure which shows the frequency characteristic of the complex coefficient band pass filter which used the complex coefficient band pass filter. It is a block diagram which shows the arithmetic processing of the 1st complex coefficient filter part using a complex coefficient band pass filter. It is a figure which shows the circuit structure which performs the complex arithmetic process of the 1st complex coefficient filter part using a complex coefficient band pass filter. It is a figure which shows the waveform of the three-phase alternating current voltage containing an unbalance component and a 5th, 7th, 11th harmonic component. It is a figure which shows the waveform of the three-phase alternating current voltage which filtered the three-phase alternating current voltage shown in FIG. 8 with the complex coefficient filter part. It is a figure which shows an example of the phase characteristic in the pass band of a 2nd complex coefficient filter part. It is a figure which shows the circuit structure which performs the arithmetic processing of a phase difference calculation part. It is a figure which shows the circuit structure which performs the arithmetic processing of a frequency calculation part. It is a figure which shows the result of having simulated the response characteristic of the frequency detector which concerns on 1st Embodiment. It is the figure which expanded the fluctuation state of the frequency output from a frequency detector 0.3 second after the simulation start of the simulation result shown in FIG. It is a figure which shows the block configuration of 2nd Embodiment of a frequency detector. It is a figure which shows the circuit structure which performs the complex arithmetic process of the 2nd complex coefficient filter part in the frequency detector of 2nd Embodiment. It is a figure which shows the block configuration of 3rd Embodiment of the frequency detector which concerns on this invention. It is a figure which shows the multistage structure of the complex coefficient notch filter provided in the 1st complex coefficient filter part in the frequency detector of 3rd Embodiment. It is a figure which shows the frequency characteristic of the 1st complex coefficient filter part using a complex coefficient notch filter. It is a block diagram which shows the arithmetic processing of the 1st complex coefficient filter part using a complex coefficient notch filter. It is a figure which shows the circuit structure which performs the complex arithmetic process of the 1st complex coefficient filter part using a complex coefficient notch filter. It is a figure which shows the result of having simulated the response characteristic of the frequency detector which concerns on 3rd Embodiment. It is the figure which expanded the fluctuation state of the frequency output from a frequency detector 0.3 second after the simulation start of the simulation result shown in FIG.

  Hereinafter, embodiments of the present invention will be specifically described with reference to the accompanying drawings.

  FIG. 1 is a diagram showing a configuration of a grid-connected inverter device including an isolated operation detection device according to the present invention. FIG. 2 is a block diagram showing a basic configuration of processing for generating a PWM signal in the control device 6. In general, a power system includes a three-phase power system and a single-phase power system. In this embodiment, the power system B is a three-phase power system, and the grid-connected inverter device A supplies power to the power system B through a three-phase AC. This will be described as a three-phase inverter device that outputs.

  The grid-connected inverter device A converts DC power generated by the DC power source 1 into AC power by the inverter 2, removes switching noise generated by the inverter 2 by the filter 3, and adjusts the output level by the transformer 4. , An inverter having a basic configuration for outputting to the power system B via the circuit breaker 5. When the grid-connected inverter device A monitors the control device 6 that controls the output power of the inverter 2 and whether or not the grid-connected inverter device A is in a single operation state, The isolated operation detection device 7 for opening the circuit breaker 5 and disconnecting the grid interconnection inverter device A is provided.

  The control device 6 controls the AC voltage output from the grid interconnection inverter device A by feedback control in order to link the grid interconnection inverter device A to the power grid B. On the other hand, the isolated operation detection device 7 detects the frequency f of the AC voltage output from the grid connection inverter device A as described later, and the system connection inverter device A enters the single operation state using the detected value. Detect that. For the feedback control of the control device 6, a detector for detecting the AC current and the AC voltage output from the grid interconnection inverter device A is provided at an appropriate position on the output line of the grid interconnection inverter device A. In FIG. Only the AC voltage detector 8 for detecting the frequency f of the output voltage of the grid interconnection inverter A is described.

  The DC power source 1 is, for example, a power source that outputs electric energy generated by solar power generation, solar thermal power generation, wind power generation, or the like in a direct current or a battery power source such as a fuel cell. The DC power supply 1 can be applied with a power supply device that outputs any energy such as light energy, mechanical energy, thermal energy, etc., but in the following description, an example using a solar cell as the DC power supply 1 will be described. .

  Inverter 2, for example, six switching elements (IGBT (Insulated Gate Bipolar Transistor), MOSFET (Metal-Oxide-Semiconductor Field-Effect Transistor), GTO (Gate Turn-Off thyristor) and other semiconductor switching elements) are bridge-connected. This is a voltage controlled inverter composed of a full bridge circuit. A full bridge circuit is a well-known circuit in which three series circuits (arms) in which two switching elements are connected in series are connected in parallel to a pair of power supply lines, and the connection point of each arm is the output terminal. .

  The pair of switching elements of each arm is controlled to be turned on / off by a set of pulse width modulation signals (PWM signals) input from the control device 4. A set of PWM signals input to each arm is a PWM signal whose phases are inverted. Further, the three sets of PWM signals input to the three arms have the same pulse waveform, but are out of phase with each other by 2π / 3. In the inverter 2, the relative voltages at the connection points of the three arms are sequentially switched by three sets of PWM signals, whereby the DC voltage input from the DC power source 1 is converted into a three-phase AC voltage and three Output from the connecting point of the arm.

  The filter 3 removes high frequency switching noise included in the three-phase AC voltage by the switching operation of the switching element in the inverter 2. The filter 3 is composed of a low-pass filter including an L-shaped circuit of a reactor and a capacitor. The transformer 4 is constituted by a three-phase transformer, and boosts or steps down the AC voltage output from the filter 3 to a level almost the same as the system voltage of the power system B.

  The circuit breaker 5 is composed of, for example, an electromagnetic switch. The breaker 5 opens when a detection signal for detecting the isolated operation state of the grid-connected inverter device A is output from the isolated operation detection device 7 and connects the grid-connected inverter device A and the power system B. Disconnect. The circuit breaker 5 is also controlled by the control device 6, and the control device 6 disconnects the connection between the grid connection inverter device A and the power system B by the circuit breaker 5 when the operation of the grid connection inverter device A is abnormal. However, the configuration is omitted in FIG.

The control device 6 is configured by a microcomputer or an FPGA (Field-Programmable Gate Array), and performs a PWM signal generation process by digital arithmetic processing. The control device 6 generates the control targets v _uo , v _vo , and v _wo (sine wave AC voltages that have passed through the filter 3) of the output voltage of the inverter 2 as modulated waves for the U, V, and W phases, A PWM signal is generated by comparing the control targets v _uo , v _vo , and v _wo with a predetermined triangular wave v _t that is a carrier wave.

The control target v _uo of the U phase is that the output current i _u of the inverter 2 is the impedance between the inverter 2 and the power system B (mainly the impedance due to the reactor of the filter 3 and the transformer 4; hereinafter referred to as “reactor for interconnection”) This corresponds to the amplitude of the voltage obtained by vector-combining the voltage corresponding to the voltage drop due to flowing through the system voltage. Since the system voltage is controlled by the power system B, the control device 6 controls the control target v _uo by controlling the voltage of the interconnecting reactor. Since the interconnection reactor is fixed as design values of the filter 3 and the transformer 4, the voltage of the interconnection reactor is controlled by the output current i _u of the inverter 2. Therefore, the control device 6 controls the control target v _uo by substantially controlling the inverter 2 output current i _u . The same applies to the control targets v _vo and v _wo for the V phase and the W phase.

The bus voltage target value generation unit 6a, the reactive power target value generation unit 6b, the reactive power calculation unit 6c, the uvw-dq conversion unit 6d, the PI compensation units 6e, 6f, 6g, and 6h in the control device 6 illustrated in FIG. The interference units 6i and 6j and the dq-uvw conversion unit 6k are processing blocks for generating output voltage control targets v _uo , v _vo and v _wo by a power major loop. A block surrounded by a dotted line (processing portion by the uvw-dq conversion unit 6d, the PI compensation units 6g and 6h, the non-interacting units 6i and 6j, and the dq-uvw conversion unit 6k) is a portion constituting a current minor loop. The PWM signal generator 6m is a processing block that compares the control targets v _uo , v _vo , and v _wo with the triangular wave v _t to generate a PWM signal. Note that a processing block for adding a system voltage counter component is provided at the front stage or the rear stage of the dq-uvw conversion unit 6k, but the processing block is omitted in FIG.

The control device 6 generates control targets I _do and I _qo for the output current of the inverter 2 in a dq rotating coordinate system (a coordinate system rotating at the frequency of the power system B). That is, the control device 6 sets the reference value V _ref of the direct-current voltage (hereinafter referred to as “bus voltage”) input to the inverter 2 by the bus voltage target value generation unit 6a, and the bus voltage reference value V _ref A deviation ΔV _dc = V _ref −V _dc of the bus voltage V _dc actually measured by the DC voltmeter 8 is obtained, and a predetermined PI compensation calculation is performed on the deviation ΔV _dc by the PI compensation unit 6e. The d-axis component I _dref of the output current control reference is set.

In addition, the control device 6 receives the reactive power target value Q _o generated by the reactive power target value generation unit 6b (Q _o = 0 when operating with a power factor of 1) and is input from the isolated operation detection device 7 described later. The power fluctuation value ΔQ is added to set the reactive power target value Q _o ′ = Q _o + ΔQ for detecting an isolated operation. Reactive power variation value ΔQ is the variation value for varying the reactive power Q _o output from the inverter 2 to detect that the system interconnection inverter device A shifts to the islanding state actively.

The control device 6 uses the output voltages v _u , v _v , v _{w of the} inverter 2 measured by the AC voltmeter 9 and the output currents i _u , i _v , i _w of the inverter 2 measured by the AC ammeter 10. calculating the reactive power Q _r output from the inverter 2 by the reactive power calculation unit 6c Te 'deviation ΔQ of the reactive power calculated value Q _r for = Q _o' the reactive power target value Q _o for islanding detection -Q _r Ask for. Then, the PI compensation unit 6f performs a predetermined PI compensation operation on the deviation ΔQ to set the control axis q-axis component I _qref of the output current of the inverter 2 in the dq rotational coordinate system.

The control device 6 uses the uvw-dq converter 6d to output the output currents i _u , i _v , i _w of the inverter 2 detected by the AC ammeter 10,
I _d = √ (2/3) · (i _u · cos (θ) + _iv · cos (θ-2π / 3) + i _w · cos (θ-4π / 3))
I _q = √ (2/3) · (−i _u · sin (θ) −i _v · sin (θ−2π / 3) −i _w · sin (θ−4π / 3))
However, θ = 2πf _s · t (f _s : system frequency)
Of converting the d-axis component I _d and a q-axis component I _q of dq rotating coordinate system by uvw-dq coordinate transformation equation, the control reference I _dref, dq-axis component in the dq rotating coordinate system of the measured value for I _qref I _{d, Iq} deviations ΔI _d = I _dref −I _d and ΔI _q = I _qref −I _q are calculated, respectively.

Controller 6, as well as a predetermined PI compensation calculation on the deviation [Delta] I _d by PI compensator 6 g, calculating the amount of interference to the q-axis component I _q in the dq rotating coordinate system of the actual measurement value is multiplied by the impedance component ωL of the filter 3 Then, the calculated value is added to the PI compensation calculated value of the deviation ΔI _d to set the control target d-axis component I _do of the output current of the inverter 2 in the dq rotational coordinate system. In addition, the control device 6 performs a predetermined PI compensation calculation on the deviation ΔI _q by the PI compensation unit 6h, and multiplies the d-axis component I _d in the dq rotation coordinate system of the actually measured value by the impedance component ωL of the filter 3 to generate an interference amount. And the calculated value is subtracted from the PI compensation calculated value of the deviation ΔI _q to set the q-axis component I _qo of the control target of the output current of the inverter 2 in the dq rotating coordinate system.

Then, the control device 6 adds a system voltage counter component (not _shown) to the control targets I _do and I _qo , respectively, and dq axis components V _do and V _qo of the control target of the output voltage of the inverter 2 in the dq rotating coordinate system. And the control targets V _do and V _qo are calculated by the dq-uvw converter 6k.
v _uo = [√ (2/3)] · [V _do · cos (θ) −V _qo · sin (θ)]
v _vo = [√ (2/3)] · [V _do · cos (θ−2π / 3) −V _qo · sin (θ−2π / 3)]
v _wo = [√ (2/3)] · [V _do · cos (θ−4π / 3) −V _qo · sin (θ−4π / 3)]
The control targets v _uo , v _vo , and v _wo of the U, V, and W phases are generated by converting the three-phase voltages in the stationary coordinate system using the dq-uvw coordinate conversion formula.

Then, the control device 6 compares the levels of the control targets v _uo , v _vo , and v _{wo with} the level of the triangular wave v _{t in} the PWM signal generation unit 6m, and generates a pulse signal having a level according to the comparison result. PWM signals for U, V, and W phases are generated. The PWM signal for each phase is input to the switching element of each arm of the inverter 2.

  The isolated operation detection device 7 detects whether or not the grid interconnection inverter device A is in an isolated operation state by the QC mode frequency shift method. As described above, in the QC mode frequency shift method, when the output frequency of the grid connection inverter A rises with respect to the reference frequency (system frequency), the phase of the output current of the grid connection inverter A is advanced and positive feedback is performed. When the output frequency increases and the output frequency of the grid connection inverter A decreases with respect to the reference frequency (system frequency), the output current is decreased by positive feedback to delay the phase of the output current of the grid connection inverter A This is a method of amplifying the fluctuation amount of the output frequency by accelerating the motor and detecting the isolated operation based on the fluctuation amount.

Since the phase of the output current of the grid interconnection inverter A is controlled by controlling the reactive energy, the isolated operation detection device 7 detects the rate of change in the frequency of the alternating current output from the grid interconnection inverter A. Based on the detected value, a fluctuation value ΔQ for changing the reactive power target value Q _o is set and fed back to the control device 6.

  Therefore, the isolated operation detection device 7 includes a reactive power fluctuation controller 71 including a frequency change rate calculating unit 71A and a reactive power fluctuation value generating unit 71B, a frequency detector 72 including a disturbance removing unit 72A and a frequency extracting unit 72B, and an isolated operation. A detector 73 is provided. The isolated operation detection device 7 is constituted by a microcomputer or FPGA, and processes each unit by digital arithmetic processing.

The reactive power fluctuation controller 71 calculates the change rate (df / dt) of the frequency f detected by the frequency detector 72 by the frequency change rate calculation unit 71A, and the reactive power fluctuation value generation unit 71B calculates the change rate (df The reactive power fluctuation value ΔQ for changing the reactive power target value Q _o according to the magnitude and polarity of / dt) is generated. The reactive power fluctuation value generation unit 71B generates a reactive power fluctuation value ΔQ using a relational expression between a preset frequency change rate (df / dt) and the reactive power fluctuation value ΔQ, and controls the reactive power fluctuation value ΔQ. Input to device 6.

In the control device 6, as described above, the process of adding the fluctuation value ΔQ input from the reactive power fluctuation value generation unit 71B to the reactive power target value Q _o generated by the reactive power target value generation unit 6b is performed. As a result, the reactive power target value Q _o ′ fluctuates in accordance with the frequency change rate (df / dt), and the control target of the output current of the inverter 2 in the dq rotating coordinate system in accordance with the fluctuation of the reactive power target value Q _o ′. Since the q-axis component I _qo varies, the phases of the three-phase voltage control targets v _uo , v _vo , and v _wo obtained by dq-uvw conversion of the dq-axis components I _do and I _qo vary.

When the grid interconnection inverter A enters the single operation state, the phase fluctuation amount of the control targets v _uo , v _vo , v _wo of the three-phase voltage increases, and the frequencies of the control targets v _uo , v _vo , v _wo are increased. Since this greatly deviates from the system frequency, the isolated operation detector 73 of the isolated operation detection device 7 detects the state and the circuit breaker inverter A is disconnected by the circuit breaker 5 as described later.

  The frequency detector 72 detects the frequency f of the output voltage of the grid interconnection inverter device A detected by the AC voltage detector 8. The isolated operation detection device 7 according to the present embodiment is characterized by the configuration of the frequency detector 72. Hereinafter, the configuration of the frequency detector 72 will be described in detail.

  FIG. 2 is a diagram illustrating a block configuration of the first embodiment of the frequency detector 72.

The frequency detector 72 is an unbalanced component or a harmonic component included in the three-phase AC voltage (phase voltages of U, V, and W) v _u , v _v , and v _w detected by the AC voltage detector 8. Etc. (components that cause disturbance in frequency detection processing) 72 A, and an AC signal output from the disturbance removal unit 721 using the phase characteristics of a band-pass type complex coefficient filter (complex coefficient bandpass filter) And a frequency extraction unit 72B that calculates the frequency of (the fundamental wave component of the three-phase AC voltage).

The disturbance removing unit 72A is a two-phase AC voltage vα, vβ orthogonal to the three-phase AC voltages v _u , v _v , v _w (instantaneous values input at a predetermined sampling period) input from the AC voltage detector 8. converting the three-phase / two-phase conversion unit 721, the three-phase / two-phase conversion unit 721 the two-phase AC voltages vα output from, (a component of the frequency -f _s) imbalance component included in vβ a predetermined order of Complex coefficient band-pass filter that removes harmonic components (mainly harmonic components such as fifth-order harmonic components (−5 f _s ), seventh-order harmonic components (+7 f _s ), and eleventh-order harmonic components (−11 f _s )) And a normalization unit 723 that normalizes the two-phase AC voltages v _r and v _j not including the unbalanced component and the harmonic component output from the first complex coefficient filter unit 722. And including.

The three-phase / two-phase conversion unit 721 performs a process of converting the three-phase AC voltages v _u , v _v , v _w input from the AC voltage detector 8 into two-phase AC voltages vα, vβ. As shown in FIG. 4A, the horizontal u-axis as a stationary coordinate system is used as a reference axis of phase θ, and the v-axis and w-axis are opposite to each other at an angle of ± 2π / 3 with respect to the u-axis. A uvw coordinate system arranged clockwise and an αβ coordinate system (stationary orthogonal coordinate system) arranged with an α axis along the u axis and a β axis orthogonal to the α axis are provided, and in these coordinate systems, the angular velocity ω Considering the voltage vector V rotating counterclockwise, the symmetric three-phase AC voltages v _u , v _v , and v _w are considered to be the u-axis component, the v-axis component, and the w-axis component of the voltage vector V, and the two-phase AC The voltages vα and vβ can be considered as an α-axis component and a β-axis component of the voltage vector V.

Therefore, the three-phase to two-phase conversion process is a process for converting the u-axis component v _u , v-axis component v _v and w-axis component v _w of the voltage vector V into an α-axis component vα and a β-axis component vβ of the voltage vector V. is there. In the following description, in order to distinguish between an AC signal and an AC signal vector, the AC signal is expressed in lower case and the AC signal vector is expressed in upper case.

The conversion equation for converting the three-phase AC voltages v _u , v _v and v _w into the two-phase AC voltages vα and vβ is as follows:
It is.

The three-phase / two-phase converter 721 converts the three-phase AC voltages v _u , v _v , v _w input from the AC voltage detector 8 into a two-phase AC by performing the calculations of the equations (1) and (2). The voltages vα and vβ are converted.

In general, the three-phase AC voltages v _u , v _v , and v _w detected by the AC voltage detector 8 are not balanced components and odd-order components such as third-order, fifth-order, seventh-order, and eleventh-order. It is an asymmetric three-phase AC voltage that includes a harmonic component (see the frequency component in FIG. 5). Therefore, the three-phase / two-phase conversion unit 721 outputs two-phase AC voltages vα ′ and vβ ′ including components obtained by three-phase to two-phase conversion for these components.

The fundamental wave components v _su , v _sv , and v _sw of the three-phase AC voltages v _u , v _v , and v _w are components of u, v, and w in the uvw coordinate system of the fundamental wave voltage vector V _s. Defined. On the other hand, the unbalanced components v _su ′, v _sv ′, v _sw ′ of the three-phase AC voltages v _u , v _v , v _w are represented by the u-axis with respect to the uvw coordinate system, as shown in FIG. In the uwv coordinate system in which the arrangement order of the v-axis and the w-axis is reversed, the components are defined as components in the u, v, and w axis directions of the voltage vector V _s ′ of the unbalanced component.

The fundamental wave components v _su , v _sv , v _sw
v _su = A _s · cos (ω _s · t) (3A)
v _sv = A _s · cos (ω _s · t -2π / 3) (3B)
v _sw = A _s · cos (ω _s · t -4π / 3) (3B)
However, if A _{s is} the amplitude of the fundamental wave component, the unbalanced components v _su ′, v _sv ′, and v _sw ′ are
v _su '= A _s ' · cos (ω _s · t) (4A)
v _sv '= A _s ' · cos (ω _s · t -4π / 3) (4B)
v _sw '= A _s ' · cos (ω _s · t -2π / 3) (4C)
However, A _s ′ is represented by the amplitude of the unbalanced component.

Substituting Equations (3A) to (3C) into Equations (1) and (2) to obtain the two-phase AC voltages v _s α and v _s β of the fundamental component,
v _s α = √ (3/2) · A _s · cos (ω _s · t) (5)
v _s β = √ (3/2) · A _s · sin (ω _s · t) (6)
It becomes. Further, when the equations (4A) to (4C) are substituted into the equations (1) and (2) to obtain the two-phase AC voltages v _s α ′ and v _s β ′ as unbalanced components,
v _s α ′ = √ (3/2) · A _s ' · cos (ω _s · t)
v _s β ′ = − √ (3/2) · A _s ′ · sin (ω _s · t)
It becomes. Since cos (ω _s · t) = cos (−ω _s · t) and −sin (ω _s · t) = sin (−ω _s · t), the two-phase AC voltage vs _{s of the} unbalanced component α ′, v _s β ′ is
v _s α ′ = √ (3/2) · A _s ' · cos (−ω _s · t) (7)
v _s β ′ = √ (3/2) · A _s ′ · sin (−ω _s · t) (8)
It becomes.

Comparing Equations (7) and (8) with Equations (5) and (6), the angular frequency of the fundamental component is “ω _s ”, whereas the angular frequency of the unbalanced component is “−ω. _The difference is that “ _s ”. When the angular frequency ω _{s is set} to “positive frequency”, the angular frequency −ω _s of the unbalanced component becomes “negative frequency”, and thus the unbalanced components v _su ′, v _sv ′, and v _sw ′ are changed to three-phase two. The two-phase AC voltages v _s α ′ and v _s β ′ obtained by phase conversion can be said to be voltages having a negative frequency.

5, displays a fundamental component in the position of the frequency "f _s" of the positive frequency domain, the frequency "-f _s negative frequency domain unbalanced component frequencies of the unbalanced components as" -f _s """ Indicates the relationship of the above frequencies. In FIG. 5, only the fifth-order, seventh-order, and eleventh-order harmonic components are depicted as the harmonic components that affect the frequency detection. The harmonic component of an integer multiple of 3 does not appear in the line voltage, and even the phase voltage is removed by the Δ-connected transformer, and the odd-order harmonic component larger than the 11th order has a low level and can be ignored. It is not described in 5.

The fifth order, seventh order, and eleventh order harmonic components v _nu , v _nv , v _nw (where n is the order, n = 5, 7, 11) of the fundamental wave components v _su , v _sv , v _sw are
_{v nu = A n · cos (} n · ω s · t) ... (9)
v _nv = A _n · cos (n · ω _s · t−n · 2π / 3) (10)
v _nw = A _n · cos (n · ω _s · t−n · 4π / 3) (11)
Therefore, by substituting the equations (9) to (11) into the equations (1) and (2), the two-phase AC voltage (V ₅ α, V ₅ beta), when obtaining the _{(V 7 α, V 7 β} ), (V 11 α, V 11 β),
v ₅ α = √ (3/2) · A ₅ · cos (−5ω _s · t) (12)
v ₅ β = √ (3/2) · A ₅ · sin (−5ω _s · t) (13)
v ₇ α = √ (3/2) · A ₇ · cos (7ω _s · t) (14)
v ₇ β = √ (3/2) · A ₇ · sin (7ω _s · t) (15)
v ₁₁ α = √ (3/2) · A ₁₁ · cos (−11ω _s · t) (16)
v ₁₁ β = √ (3/2) · A ₁₁ · sin (−11ω _s · t) (17)
It becomes.

The unbalanced component has a negative frequency because the phase order of the unbalanced component (uvw is clockwise) is opposite to the phase order of the fundamental component (uvw is counterclockwise). is there. Accordingly, two-phase AC voltages v _n α, v _n β (subscript n is a subscript of three-phase two-phase conversion of an n-order harmonic component obtained by multiplying the frequency f _s of the fundamental wave component by n (n: an integer of 2 or more). If the angular frequency ω _{n of the} order (the same applies hereinafter) has the same sign as the two-phase AC voltages v _s α and v _s β of the fundamental component, the frequency f _n of the n-order harmonic component is a positive frequency Thus, the phase order of the n-order harmonic components v _nu , v _nv , and v _nw is the same as the phase order of the fundamental wave components v _su , v _sv , and v _sw . Conversely, when the angular frequency ω _n of the two-phase AC voltages v _n α and v _n β has a sign opposite to that of the two-phase AC voltages v _s α and v _s β of the fundamental wave component, the n-order harmonic component The frequency f _n is a negative frequency, and the phase order of the n-th harmonic components v _nu , v _nv , v _nw is opposite to the phase order of the fundamental wave components v _su , v _sv , v _sw .

From equations (12) to (17), the fifth and eleventh harmonic components have negative frequencies, and the seventh harmonic component has a positive frequency. order harmonic component and the 11th order harmonic component displays each negative frequency "-5F _s" in the frequency domain to the position of "-11F _s", the fifth harmonic component and the 11th order harmonic component of the opposite phase Are respectively displayed at the positions of the frequencies “5f _s ” and “11 f _s ” in the positive frequency region. Further, the positive-phase seventh harmonic component is displayed at the position of the frequency “7f _s ” in the positive frequency region, and the negative-phase seventh harmonic component is displayed at the position of the frequency “−7 f _s ” in the negative frequency region. Is displayed.

Therefore, from the three-phase / two-phase conversion unit 721, the fundamental wave component, the unbalanced component, the fifth order, the seventh order, and the eleventh order represented by the expressions (5) to (8) and (12) to (17). Two-phase AC voltages (v _s α, v _s β), (v _s α ′, v _s β ′), (v _n α, v _n β) (n = 5, 7, 11) Including two-phase AC voltages vα ′ and vβ ′ are output.

The first complex coefficient filter unit 722 is composed of a first-order IIR bandpass filter whose transfer function H (z) expressed by z-transform expression is expressed by the following equation (18), and has the frequency characteristics shown in FIG. ing. In equation (18), f _d [Hz] in the complex coefficient a ₁ is a normalized frequency obtained by normalizing the center frequency f _{0 of the} passband with the sampling frequency. Ω _d [rad / s] is a normalized angular frequency. For example, when the sampling frequency is “f _sr ” and the center frequency f ₀ is set to the system frequency f _s , f _d is f _s / f _sr , and Ω _d is 2π · f _d = 2π · (f _s / f _sr ). It becomes. The normalized angular frequency Ω _d is −π <Ω _d <π. R is a parameter (0 <r <1) that determines the bandwidth of the passband.

FIG. 6 is a block diagram showing a processing circuit for performing the arithmetic processing of the equation (18). As shown in the figure, the first complex coefficient filter unit 722 is configured by a circuit in which the denominator arithmetic processing of the equation (18) is configured by a feedback circuit, and the output of the feedback circuit is multiplied by a numerator coefficient b _0. The

In the block diagram shown in FIG. 6, u [k] (k: index number representing discrete time) is input data, x [k] is state data of the first complex coefficient filter unit 722, and y [k] is the first complex. This is output data of the coefficient filter unit 722. Between input data u [k], state data x [k] and output data y [k]
x [k] = r · exp (j · Ω _d ) · x [k−1] + u [k] (19)
y [k] = (1-r) .x [k] (20)
Is established.

Since the first complex coefficient filter unit 722 is composed of a complex coefficient bandpass filter, the state is independent of whether the input data u [k] is complex data or real data (data whose imaginary part of complex data is “0”). Data x [k] and output data y [k] are complex signal data. Accordingly, the input data u [k], state data x [k] and the output data y [k], respectively _{u [k] = u r [} k] + j · u j [k], x [k] = x r [ k] + j · x _j [k], y [k] = y _r [k] + j · y _j [k] complex data, complex coefficient a ₁ is expressed as a ₁ = r · exp (j · Ω _d ) = a Substituting into the equations (19) and (20) as _r + j · a _j = r · cos (Ω _d ) + j · r · sin (Ω _d ), and dividing into the relational expression of the real part and the imaginary part,
_{x r [k] = r ·} cos (Ω d) · x r [k-1] -r · sin (Ω d) · x j [k-1] + u r [k] ... (21)
x _j [k] = r · cos (Ω _d ) · x _j [k−1] + r · sin (Ω _d ) · x _r [k−1] + u _j [k] (22)
y _r [k] = (1−r) · x _r [k] (23)
y _j [k] = (1−r) · x _j [k] (24)
It becomes.

FIG. 7 is a diagram showing a circuit configuration for performing the complex arithmetic processing of the first complex coefficient filter unit 722 based on the equations (21) to (24). In the figure, coefficient a _r and coefficient a _j are the real part and imaginary part of complex coefficient a ₁ = r · exp (j · Ω _d ), respectively, and a _r = r · cos (Ω _d ), a _j = r · sin (Ω _d ).

As shown in the figure, the first complex coefficient filter unit 722 includes six multipliers 722a to 722f, two adders 722g and 722h, and two delay circuits 722i and 722j. The delay circuit 722i is a circuit that generates a real part x _r [k-1] of state data, and the delay circuit 722j is a circuit that generates an imaginary part x _j [k-1] of state data. Multipliers 722a and 722b are arithmetic units for calculating the first and second terms (including a negative sign) of equation (21), respectively, and adder 722g is the first and second terms of equation (21). And an arithmetic unit for adding the third term. Accordingly, the real part x _r [k] of the state data shown by the equation (21) is output from the adder 722g.

On the other hand, the multipliers 722c and 722d are arithmetic units for calculating the first term and the second term of the equation (22), respectively, and the adder 722h is the first term, the second term, and the third term of the equation (22). An arithmetic unit to add. Therefore, the imaginary part x _j [k] of the state data indicated by the equation (22) is output from the adder 722h. Multipliers 722e and 722f are arithmetic units for calculating the expressions (23) and (24), respectively.

In the present embodiment, the three-phase / two-phase conversion unit 721 is provided to convert the three-phase AC voltages v _u , v _v , v _w into the two-phase AC voltages vα, vβ that are orthogonal to each other. Since vα and vβ can respectively correspond to the real part and imaginary part of the complex data u _r + j · u _j , the sampling data of the two-phase AC voltage vα is used as the real part u _r [k] of the input data. The sampling data of the two-phase AC voltage vβ is input to the adder 722h as the imaginary part u _j [k] of the input data.

Every time sampling data of the two-phase AC voltage vα is input to the first complex coefficient filter unit 722, the delay circuit 722i, the multipliers 722a, 722b, and 722e and the adder 722g perform the calculations of the expressions (21) and (23). As a result, the multiplier 722e outputs output data y _r [k] of only the two-phase AC voltage v _s α obtained by three-phase to two-phase conversion of the fundamental wave component represented by the equation (5). . Further, every time sampling data of the two-phase AC voltage vβ is input to the first complex coefficient filter unit 722, the delay circuit 722j, the multipliers 722c, 722d, and 722f, and the adder 722h use the expressions (22) and (24). As a result, the multiplier 722f outputs the output data y _j [k] of only the two-phase AC voltage v _s β obtained by three-phase to two-phase conversion of the fundamental wave component expressed by the equation (6). Is done.

When the bandpass filter is configured by a real coefficient second-order IIR filter, the transfer function H (z) (z = exp (j · ω)) of the second-order IIR filter is
H (z) = (1-r ² +2 (r-1) · r · cos (Ω _d ) · z ^-1 ) / (1-2r · cos (Ω _d ) · z ^-1 + r ² · z ^-2 )
It is represented by When the amplitude characteristic M (ω) of the transfer function H (z) is obtained, a pole appears at ω that satisfies M (ω) = (1-2r · cos (Ω _d ± ω) + r ² ) = 0. The next IIR filter has the property of passing the frequency of its pole. If r≈1, the normalized frequency f _d that passes through the second-order IIR filter is f _d = ± Ω _d / 2π from cos (Ω _d ± ω) ≈1, so the normalized angular frequency Ω _d is basically used. The real coefficient second-order IIR filter set to the angular frequency of the wave component also allows an unbalanced component (a component of −f _s = −Ω _d · f _sr / 2π) to pass through. That is, in the frequency characteristic shown in FIG. 5, the second-order IIR filter also passes an unbalanced component appearing in “−f _s ”.

On the other hand, when the amplitude characteristic M (ω) of the transfer function H (z) shown in the equation (18) is obtained, M (ω) = (1-r) / √ {1-2r · cos (Ω _d −ω) + r ² }, And a pole appears only in ω that satisfies (1-2r · cos (Ω _d −ω) + r ² ) = 0, so that the normalized angular frequency Ω _{d is changed} to the angular frequency (ω = Ω _d ) of the fundamental wave component. The set complex coefficient first-order IIR filter has the frequency characteristics shown in FIG. Therefore, the first-order IIR filter having a complex coefficient passes only the fundamental wave component (the component of f _s = Ω _d · f _sr / 2π) and does not pass the unbalanced component and the harmonic component.

FIG. 8 is a diagram illustrating waveforms of three-phase AC voltages v _u , v _v , and v _w including 5% of unbalanced components and fifth, seventh, and eleventh harmonic components. 9 shows a first example in which the three-phase AC voltages v _u , v _v and v _w shown in FIG. 8 are set to the three-phase / two-phase converter 721 and the center frequency f _{0 is set} to the system frequency f _s (60 [Hz]). 3 shows waveforms simulating three-phase AC voltages v _u ′, v _v ′, and v _w ′ obtained by inversely converting a two-phase AC voltage into a three-phase AC voltage after filtering by the complex coefficient filter unit 722. FIG.

Since the first complex coefficient filter unit 722 in which the center frequency f ₀ is set to the system frequency f _s has the frequency characteristics shown in FIG. 5, as shown in FIG. 9, the three-phase AC voltages v _u , v _v , v _It can be seen that the unbalanced component and the fifth, seventh, and eleventh harmonic components included in _w can be suitably removed by the first complex coefficient filter unit 722.

In FIG. 3, in order to distinguish the two-phase AC voltage output from the first complex coefficient filter unit 722 from the output data y _r [k], y _j [k] from the two-phase AC voltages vα and vβ, respectively. “V _r ” and “v _j ” are indicated.

The two-phase AC voltages v _r and v _j output from the first complex coefficient filter unit 722 and expressed by the equations (5) and (6) have a phase angle φ of the U-phase voltage v _u as “0”. In this case, when the phase of the U-phase voltage v _u of the three-phase power system B is shifted and the phase φ ≠ 0, the two-phase AC voltages v _r and v output from the first complex coefficient filter unit 722 are used. _j is v _r = A _s · cos (ω _s · t + φ) and v _j = A _s · sin (ω _s · t + φ).

The normalization processing unit 723 performs arithmetic processing for normalizing the levels of the two-phase AC voltages v _r and v _j output from the first complex coefficient filter unit 722 to “1”. The two-phase AC voltages v _r and v _j output from the first complex coefficient filter unit 722 are sine waves and cosine waves having the same amplitude, and the amplitude is obtained by calculating √ (v _r ² + v _j ² ). Therefore, in the normalization processing unit 723, y _r [k] / √ (y _r [k] ² + y _j [k] ² ) and y _{j for} the output data y _r [k] and y _j [k], respectively. A normalization process of the two-phase AC voltages v _r and v _j is performed by performing an arithmetic process of [k] / √ (y _r [k] ² + y _j [k] ² ). Therefore, the normalization processing unit 723 outputs data of signals represented by v _r ′ = cos (θ _s ) (θ _s = ω _s · t) and v _j ′ = sin (θ _s ).

The frequency extraction unit 72B includes a second complex coefficient filter unit 724 including a complex coefficient bandpass filter having a linear phase characteristic in the pass band, and a voltage vector V _in = exp (j input to the second complex coefficient filter unit 724. · ω _s · t) the phase difference calculating section for calculating a phase difference [psi in the second complex coefficient filter unit 724 voltage vector output from the _{V out = exp (j · ω} s · t + ψ) both voltage vector using the Toko including the 725, the frequency calculator 726 calculates the voltage vector V _in the frequency _{f (= (ω s / 2π} ) the phase difference ψ is calculated by the phase difference calculation unit 725 using a predetermined arithmetic expression, a.

The complex coefficient bandpass filter used in the second complex coefficient filter unit 724 has the same center frequency f ₀ as the complex coefficient bandpass filter used in the first complex coefficient filter unit 722. Further, the complex coefficient band pass filter has a characteristic in which the phase characteristic (phase characteristic with respect to frequency) in the pass band is expressed by linear approximation.

FIG. 10 is a diagram illustrating an example of phase characteristics in the pass band of the complex coefficient bandpass filter used in the second complex coefficient filter unit 724. This figure shows the change in phase ψ [deg] at f ₀ ± 5 [Hz] when the center frequency f _{0 is set} to the system frequency f _s = 60 [Hz].

  According to FIG. 10, the change of the phase ψ with respect to the frequency f changes linearly, and the relational expression of f = p · ψ + q is established between the phase ψ and the frequency f. Since f = 60 [Hz] and ψ = 0 [deg] and f = 57.25 [Hz] and ψ = 0.2 [deg], p = from 60 = q and 57.25 = 0.2 · p + q. ≈−13.75, and in the example of FIG. 10, the arithmetic expression for obtaining the frequency f from the phase characteristic and the phase ψ is f = −13.75 · f + 60.

  In the present embodiment, the complex coefficient bandpass filter used in the second complex coefficient filter unit 724 is the same as the complex coefficient bandpass filter used in the first complex coefficient filter unit 722. Therefore, the specific arithmetic circuit is the same as the circuit shown in FIG. 7, and detailed description thereof is omitted here.

The phase difference calculation unit 725 calculates the phase difference ψ using the voltage vector V _in input to the second complex coefficient filter unit 724 and the voltage vector V _out output from the second complex coefficient filter unit 724. The conjugate voltage vector V _in ^* of the voltage vector V _in = exp (j · ω _s · t) is V _in ^* = exp (−j · ω _s · t), and this voltage vector V _in ^* and the output voltage When the vector V _out = exp [j · (ω _s · t + ψ)] is multiplied, V _in ^* · V _out = exp (−j · ω _s · t + j · ω _s · t + j · ψ) = exp (j A vector of phase difference ψ is obtained from ψ). When the voltage vector V _in ^* = exp (−j · ω · t) = R _in −j · X _in and the voltage vector V _out = exp [j · (ω _s · t + ψ)] = R _out + j · X _out , Since V _in ^* · V _out = (R _in · R _out + X _in · X _out ) + j (R _in · X _out −X _in · R _out ) = cos (ψ) + j sin (ψ), sin (ψ) = (R _in · X _out -X _in · R _out ), the phase difference calculation unit 725 calculates the phase difference ψ by performing an operation of sin ^-1 (R _in · X _out -X _in · R _out ). calculate.

From cos (ψ) = (R _in · R _out + X _in · X _out ), the phase difference ψ is calculated by performing a calculation process of cos ^-1 (R _in · R _out + X _in · X _out ). From tan (ψ) = (R _in · X _out -X _in · R _out ) / (R _in · R _out + X _in · X _out ), tan ^-1 [(R _in · X _out -X _in · R _out ) / (R _in · R _out + X _in · X _out )] may be calculated to calculate the phase difference ψ.

  FIG. 11 is a diagram illustrating a circuit configuration for performing the arithmetic processing of the phase difference calculation unit 725.

The phase difference calculation unit 725 the real part R _in the imaginary part X _in the voltage vector V _in input to the second complex coefficient filter unit 724, a real voltage vector V _out output from the second complex coefficient filter unit 14 Two data of the part _Rout and the imaginary part _Xout are input. The real part R _in the voltage vector V _in is the voltage data output from the normalization processing unit _{723 v r '= cos (ω} s · t), the imaginary part X _in an output from the normalization processing unit 723 The voltage v _j ′ = sin (ω _s · t). On the other hand, the real part R _out of the voltage vector V _out is data of the voltage y _r = cos (ω _s · t + ψ) output from the second complex coefficient filter unit 724, and the imaginary part X _out is the second complex coefficient filter unit. Data of voltage y _j = sin (ω _s · t + ψ) output from 724. Therefore, if the equation for the phase difference ψ is v _r ′ · y _j −v _j ′ · y _r = Ψ,
ψ = sin ⁻¹ (Ψ) (25)
It becomes.

The phase difference calculation unit 725 includes two multipliers 725a and 725b, one adder 725c, and an inverse sine value calculator 725d. The multiplier 725a multiplies v _r '· y _j in Ψ, and the multiplier 725b multiplies v _j ' · y _r in Ψ. Further, the adder 725c calculates the argument Ψ of the inverse trigonometric function by subtracting the multiplication result of the multiplier 725b from the multiplication result of the multiplier 725a. Then, the inverse sine value calculator 725d calculates sin ⁻¹ (Ψ) for the argument Ψ output from the adder 725c, and calculates the phase difference ψ.

  The frequency calculation unit 726 calculates a predetermined calculation formula f = p · ψ + q (for calculating the frequency f based on the phase characteristic of the second complex coefficient filter unit 724 with respect to the phase difference ψ input from the phase difference calculation unit 725. In the example of FIG. 10, the frequency f is calculated by executing P = −13.5, q = 60). Therefore, the frequency calculation unit 726 includes a multiplier 726a and an adder 726b as shown in FIG. The multiplier 726a multiplies the phase difference ψ input from the phase difference calculation unit 15 by the coefficient p, and the adder 726b adds the coefficient q to the multiplication result and outputs the phase difference ψ.

  Note that a table of the frequency f and the phase difference ψ satisfying the relationship of f = p · ψ + q for a predetermined frequency f range is stored, and the phase difference ψ output from the phase difference calculation unit 725 is used as the table. The corresponding frequency f may be obtained.

FIG. 13 shows the result of simulating the response characteristic of the frequency f detected by the frequency detector 72 according to the first embodiment. Specifically, the simulation is started when the frequency of the three-phase AC voltage of the power system is stable at the system frequency f _s = 60 [Hz], and the frequency f _s is instantaneously set 0.2 seconds after the simulation starts. The response characteristic of the frequency detector 72 when the frequency is changed to 60.3 [Hz] is shown. FIG. 14 is an enlarged view of the fluctuation state of the frequency f output from the frequency detector 72 after 0.3 seconds from the start of the simulation. The content of unbalanced components and fifth, seventh, and eleventh harmonic components included in the input three-phase AC voltages v _u , v _v , and v _w is 5%, and the sampling frequency is the system frequency f _s. The high frequency is several hundred times. The center frequency f ₀ of the pass band of the first complex coefficient filter unit 722 is set to the system frequency f _s = 60 [Hz].

As shown in FIG. 13, when the frequency f _{s of the} power system is instantaneously increased from 60.0 [Hz] to 60.3 [Hz] 0.2 seconds after the start of the simulation, the frequency detector 72 outputs the signal. The frequency f is suddenly changed in a pulse shape to follow the sudden change in the frequency f _s of the power system, but is about 60.3 [Hz] in about 0.05 seconds after the sudden frequency change (time 0.2 seconds). It turns out that it converges. Further, when 0.1 second has elapsed since the sudden frequency change (time 0.3 second), the ripple of the frequency f output from the frequency detector 72 is ± 0.01 [Hz] as shown in FIG. Therefore, the frequency detector 72 changes the output frequency f when the frequency suddenly changes (time 0), since the fluctuation range is about 0.016% with respect to the frequency f _s = 60.3 [Hz] of the power system after the change. Within 2 seconds from 0.1 second), the frequency f _s after the change of the power system can be set.

Returning to FIG. 1, the isolated operation detector 73 calculates a fluctuation amount Δf (= f−f _S ) of the frequency f detected by the frequency detector 72 from the system frequency f _S , and the fluctuation amount Δf is set in advance. It is determined whether the threshold value f _th is exceeded. When the fluctuation amount Δf exceeds the threshold f _th , the isolated operation detector 73 outputs a disconnect signal to the circuit breaker 5 on the assumption that the grid-connected inverter device A has shifted to the isolated operation state. The circuit breaker 5 disconnects the connection between the single operation detector 73 and the three-phase electric power system B by the disconnection signal from the single operation detector 73.

  As described above, according to the first embodiment, since the frequency detector 72 is configured by the frequency detection device using the complex coefficient bandpass filter, the frequency of the output voltage of the system power inverter A is continuously detected with high accuracy. When the system power inverter A shifts to the single operation state, the system power inverter A can be disconnected from the three-phase power system B with high speed and high accuracy.

The first embodiment uses the linearity of the phase characteristic in the pass band of the complex coefficient bandpass filter of the second complex coefficient filter unit 724 to calculate the phase difference ψ between the input signal and the output signal of the complex coefficient bandpass filter. In this method, the frequency f of the input signal is obtained. The complex coefficient a ₁ of the complex coefficient bandpass filter of the second complex coefficient filter unit 724 can be changed, and the phase difference between the input signal and the output signal of the complex coefficient bandpass filter By performing control to change the center frequency f ₀ of the complex coefficient bandpass filter to coincide with the frequency f of the input signal based on ψ, the frequency f of the input signal is changed from the center frequency f ₀ of the complex coefficient bandpass filter. You may ask for.

Complex coefficient bandpass filter of the second complex coefficient filter unit 724, as shown in FIG. 10, at the center frequency f ₀ the phase difference ψ is zero, a negative phase difference ψ is a frequency range higher than the center frequency f ₀ Therefore, in the frequency region lower than the center frequency f ₀ , it has a characteristic that linearly changes so that the phase difference ψ becomes positive. Therefore, by looking at the positive and negative signs of the phase difference ψ, it can be seen whether the frequency f of the input signal of the complex coefficient bandpass filter is in a high frequency region or a low frequency region with respect to the center frequency f ₀ . Since the phase characteristic changes linearly, if ψ <0, the center frequency f ₀ is decreased while monitoring the phase difference ψ. If ψ> 0, the center frequency f ₀ is increased and ψ = 0. if, by controlling the complex coefficient a ₁ so as not to change the center frequency f _0, a frequency at which the center frequency f ₀ of the complex coefficient bandpass filter and [psi = 0, i.e., set to a frequency f of the input signal can do.

FIG. 15 is a diagram showing a block configuration of the second embodiment in which the complex coefficient a ₁ of the complex coefficient bandpass filter of the second complex coefficient filter unit 724 can be changed and the center frequency f ₀ is controlled to the frequency f of the input signal. It is. FIG. 16 is a diagram illustrating a circuit of complex arithmetic processing in the second complex coefficient filter unit 724 ′.

The frequency detector 72 ′ shown in FIG. 15 differs from the frequency detector 72 shown in FIG. 3 in the configuration of the frequency extraction unit 72B ′. Specifically, the frequency extraction unit 72B ′ is provided with a center frequency control unit 727 that changes the center frequency f ₀ based on the phase difference ψ calculated by the phase difference calculation unit 15 instead of the frequency calculation unit 726, and The complex coefficient a ₁ that determines the center frequency f ₀ of the two-coefficient filter unit 724 ′ is controlled by the normalized angular frequency Ω _d ′ input from the center frequency control unit 727. Further, the complex arithmetic processing circuit shown in FIG. 16 is different from the complex arithmetic processing circuit shown in FIG. 7 in that the real part of the complex coefficient a ₁ using the normalized angular frequency Ω _d ′ input from the center frequency control unit 727. A coefficient real part arithmetic circuit 722m for calculating the coefficient a _r of the imaginary part and a coefficient imaginary part arithmetic circuit 722k for calculating the coefficient a _j of the imaginary part are added.

In the second embodiment, the center frequency control unit 727 compares the phase difference ψ input from the phase difference calculation unit 725 with “0”, and the comparison result is any one of ψ> 0, ψ <0, and ψ = 0. Accordingly, the normalized angular frequency Ω _d of the second complex coefficient filter unit 724 ′ is changed. Since Ω _d = 2π · f _d , f _d = f ₀ / f _sr , and f _d corresponds to the center frequency f ₀ of the second complex coefficient filter unit 724 ′, the normalized angular frequency Ω _d ′ is changed. As a result, the center frequency f ₀ of the second complex coefficient filter unit 724 ′ changes.

The center frequency control unit 727 performs an operation of ψ × 180 / π on the phase difference ψ [rad] output from the phase difference calculation unit 725 to convert it into a phase difference ψ ′ [deg] in units of angles. Thereafter, (−K × ψ ′ / f _sr ) (K: predetermined gain of 0 to 1, for example, K = 0.1, f _sr : sampling frequency) is calculated to calculate the change amount ΔΩ _d. Then, the amount of change ΔΩ _d is added to Ω _ds = 2π · (f _s / f _sr ) to calculate the change value Ω _d ′. Ω _ds is a normalized angular frequency in which the center frequency f _{0 is set} to the system frequency f _s (for example, 60 [Hz]).

The phase difference ψ is given a positive or negative sign in accordance with ψ <0 or 0 ≦ ψ in the calculation processing of the expression (25). Therefore, in the addition processing of the variation ΔΩ _{d to} the normalized angular frequency Ω _ds , When ψ <0, the variation ΔΩ _d is subtracted from the normalized angular frequency Ω _ds , and when 0 ≦ ψ, the variation ΔΩ _d is added to the normalized angular frequency Ω _ds . The calculation process for changing the normalized angular frequency Ω _d ′ is merely an example, and the normalization angular frequency Ω _d ′ may be increased or decreased according to the sign of the phase difference ψ by other calculation methods. Good.

Together 'is = Ω _ds ± ΔΩ _d second complex coefficient filter 724' center normalized angular frequency is changed and set in the frequency control unit 727 Omega _d is fed back to the relative normalized angular frequency Omega _d '(Omega _d ′ × f _sr / 2π) is calculated to calculate a center frequency change value f ₀ ′ (= f _s ± Δf), and the change value f _s ′ is output as a frequency detection value f ₀ ′. .

Normalization angular frequency Omega _d 'second complex coefficient filter unit 724 is fed back', using the the normalized angular frequency Omega _d 'feedback from the center frequency control section 17 by the coefficient a real part calculating circuit 722M,
a _r = r · cos (Ω _d ') (26)
Is used to calculate the real part coefficient a _r of the complex coefficient a ₁ and the calculated value a _r is input to the multiplier 722a and the multiplier 722d.

Further, using the normalized angular frequency Ω _d ′ fed back from the center frequency control unit 17 in the coefficient imaginary part arithmetic circuit 722k,
a _j = r · sin (Ω _d ') (27)
The coefficient a _j of the imaginary part of the complex coefficient a ₁ is calculated by the following equation, and the calculated value a _j is input to the multiplier 722b and the multiplier 722c.

Therefore, if the phase difference ψ input from the phase difference calculation unit 725 is not “0”, the complex coefficient a ₁ changes in the direction in which the phase difference ψ becomes “0”. Therefore, the multiplier 722e and the multiplier 722f Is the frequency of the fundamental component of the three-phase AC voltages v _u , v _v and v _w detected by the AC voltage detector 8, that is, output data y _r of the real part of the voltage vector having the same frequency as the system frequency f _s. [k] and output data y _j [k] of the imaginary part are output.

In the second embodiment, the frequency detection value f ₀ ′ output from the center frequency control unit 727 is stable in the frequency f _{s of the} three-phase AC voltages v _u , v _v and v _w input to the frequency detector 72 ′. if the, shows the frequency f _s, if the frequency f _s fluctuates, changes as follows the variation, the three-phase AC voltage v _u that is input, v _v, the frequency f _s of the v _w It will be shown.

As described above, also in the second embodiment, the frequency f _s of the output voltage of the system power inverter A can be continuously detected with high accuracy, and when the system power inverter A shifts to the single operation state, high speed and high accuracy are achieved. Thus, the system power inverter A can be disconnected from the three-phase power system B.

  In the second embodiment, a complex coefficient bandpass filter is used for the first complex coefficient filter unit 722, but an unbalanced component, fifth order, seventh order, You may make it remove harmonic components, such as 11th order.

  FIG. 17 is a diagram illustrating a block configuration of the third embodiment using a complex coefficient notch filter as the first complex coefficient filter unit 722 of the second embodiment. FIG. 18 is a diagram illustrating an example of a multistage configuration of a complex coefficient notch filter provided in the first complex coefficient filter unit 722 '.

When the complex coefficient notch filter is used, if the frequency of the fundamental wave component of the input signal is shifted, the frequencies of the unbalanced component and the harmonic component are also shifted accordingly. Therefore, it is necessary to shift the stop frequency of the complex coefficient notch filter. Therefore, in the third embodiment, the normalized angular frequency Ω _d ′ output from the center frequency control unit 727 is also fed back to the first complex coefficient filter unit 722 ′ in the disturbance removal unit 72A ′, and the complex coefficient notch filter The complex coefficient that determines the stop frequency is changed.

As shown in FIG. 5, the AC signal of the three-phase power system B includes an unbalanced component (−f _s ) and fifth, seventh, and eleventh harmonic components (−5f) in addition to the fundamental component (+ f _s ). _s , + 7f _s , −11f _s ) are included, the first complex coefficient filter unit 722 ′ shown in FIG. 17 has a transfer function H (z) expressed by z-transform expression for each frequency shown below. A complex coefficient notch filter having a frequency characteristic shown in FIG. 19 is configured by providing complex coefficient notch filters represented by the equation (28) of FIG.

Note that Ω _d is Ω _d = 2π · (n · f _s / f _sr ), and when n = −1, a complex coefficient notch filter having an unbalanced component (−f _s ) as a stop frequency is obtained. . Further, when n = −5, n = + 7, and n = −11, a complex coefficient notch filter having a fifth-order, seventh-order, and eleventh-order harmonic components as stop frequencies is obtained. Therefore, as shown in FIG. 18, the first complex coefficient filter unit 722 ′ includes complex coefficient notch filters 722a, 722b, 722c, and 722d having −f _s , −5f _s , + 7f _s , and −11f _s as stop frequencies. The configuration is cascaded.

  The block diagram of the processing circuit that performs the arithmetic processing of the above equation (28) has the configuration shown in FIG. 20, and the circuit that performs the complex arithmetic processing of the complex coefficient filter unit 12 using the complex coefficient notch filter (NF) The configuration is as shown in FIG.

FIG. 20 is obtained by adding a circuit that subtracts data y [k] from input data u [k] and outputs the subtraction value as output data e [k] to the block diagram shown in FIG. . Further, FIG. 21 calculates the coefficient a _r having the center frequency of the frequency of the unbalanced component or the harmonic component by using the coefficient real part arithmetic circuit 722m and the coefficient imaginary part arithmetic circuit 722k with respect to the block diagram shown in FIG. The coefficient real part arithmetic circuit 722m ′ and the coefficient imaginary part arithmetic circuit 722k ′ for calculating the coefficient a _j are changed. Then, an adder 722n is added after the multiplier 722e of the real part, and the real part y _r [k] of the data y [k] is subtracted from the real part u _r [k] of the input data by the adder 722n. Thus, the real part _er [k] of the output data is output. Further, an adder 722o is added after the multiplier 722f of the imaginary part, and the adder 722o subtracts the imaginary part y _j [k] of the data y [k] from the imaginary part u _j [k] of the input data. Thus, the imaginary part e _j [k] of the output data is output.

The coefficient real number arithmetic circuit 722m ′ uses the normalized angular frequency Ω _d ′ fed back from the center frequency control unit 727,
a _r = r · cos (n · Ω _d ') (29)
Is used to calculate the real part coefficient a _r of the complex coefficient a ₁ and the calculated value a _r is input to the multiplier 722a and the multiplier 722d.

In addition, the coefficient imaginary part arithmetic circuit 722k ′ uses the normalized angular frequency Ω _d ′ fed back from the center frequency control unit 727,
a _j = r · sin (n · Ω _d ') (30)
The coefficient a _j of the imaginary part of the complex coefficient a ₁ is calculated by the following equation, and the calculated value a _j is input to the multiplier 722b and the multiplier 722c.

  Note that n represents the order of the stop frequency, n = −1 for the complex coefficient notch filter 722a shown in FIG. 18, and n = −5, +7, −11 for the complex coefficient notch filters 722b, 722c, and 722d, respectively. Become.

  The circuit shown in FIG. 21 is different from the circuit shown in FIG. 16 in that subtraction processing in adders 722n and 722o is performed in addition to the calculation contents of the coefficient real part arithmetic circuit 722m ′ and coefficient imaginary part arithmetic circuit 722k ′. Since only the added points are different, detailed description of the arithmetic processing of the circuit shown in FIG. 21 is omitted.

The frequency detector 72 'according to the third embodiment, the unbalanced component by the complex coefficient notch filter 722a~722d of (components of -f _s) and harmonics (-5f _s, + 7f _s, components -11f _s) Since the levels are respectively suppressed, it is possible to suitably prevent these components from being input to the frequency extraction unit 72B ′. Therefore, the frequency detector 72 ″ according to the third embodiment can stably and highly accurately. The frequency f _s of the three-phase voltage of the three-phase power system B can be detected.

  FIG. 22 shows the result of simulating the response characteristics of the frequency f detected by the frequency detector 72 ″ according to the third embodiment. FIG. 23 shows the frequency detector 72 ″ after 0.3 seconds from the start of the simulation. It is the figure which expanded the fluctuation state of the frequency f output from FIG. The simulation conditions are the same as the simulation results shown in FIGS.

As shown in FIG. 22, also in the case of the third embodiment, when the frequency f _{s of the} power system is instantaneously increased from 60.0 [Hz] to 60.3 [Hz] after 0.2 seconds from the start of simulation, It can be seen that the frequency f output from the frequency detector 72 ″ suddenly changes in a pulse shape, but converges to approximately 60.3 [Hz] approximately 0.05 seconds after the sudden frequency change (time 0.2 seconds). In the third embodiment, when 0.1 second has elapsed from the time of sudden frequency change (time 0.3 second), as shown in Fig. 23, the ripple of the frequency f output from the frequency detector 72 " However, it can be seen that the frequency f _s of the three-phase voltage of the three-phase power system B can be detected at a higher speed and with higher accuracy than in the first embodiment.

  As described above, according to the third embodiment, since the frequency detector 72 ″ is configured by the frequency detector using the complex coefficient notch filter, the frequency of the output voltage of the system power inverter A is set to the first and second frequencies. It can be detected at a higher speed than in the second embodiment, and when the system power inverter A shifts to the single operation state, the system power inverter A is disconnected from the three-phase power system B at a higher speed than in the first and second embodiments. Can do.

  Note that either the complex coefficient bandpass filter or the complex coefficient notch filter may be used for the first complex coefficient filter unit 722. However, it is preferable to use the complex coefficient notch filter faster than the complex coefficient bandpass filter. Accurate phase detection characteristics can be obtained. In addition, when a complex coefficient notch filter and a complex coefficient bandpass filter are combined, a synergistic effect of the characteristics of both can be expected, and a faster and more accurate phase detection characteristic can be obtained.

  As is well known, if the complex coefficient notch filter and the complex coefficient band pass filter have a multi-stage configuration, a steep filter characteristic can be obtained, and an unbalanced component and a harmonic component can be easily removed and responsive. Therefore, when mounting, a multi-stage configuration with an appropriate number of stages is preferable.

In the frequency detector 72 according to the first embodiment, the case where the first complex coefficient filter unit 722 is configured by a complex coefficient bandpass filter has been described. However, as in the third embodiment, the first complex coefficient filter unit 722 is complex. You may comprise with a coefficient notch filter. However, in this case, as described in the third embodiment, the stop frequency of the complex coefficient notch filter is set to frequencies of −f, −5f, 7f, and −11f with respect to the frequency f calculated by the frequency calculation unit 16. Need to control. Therefore, in this case, the change value of the normal angular frequency Omega _d downstream of the phase difference calculating section 15 'of the arithmetic processing unit for calculating the provided calculated change value of the normal angular frequency Omega _d in that the arithmetic processing unit' complex The first complex coefficient filter unit 722 ′ configured with a coefficient notch filter may be fed back.

Further, when the first complex coefficient filter unit 722 of the first embodiment is configured with a complex coefficient notch filter, the complex coefficient notch filter is combined with a complex coefficient band-pass filter to reduce the unbalanced component and the harmonic component. The removal effect may be enhanced. In this case, the center frequency of the complex coefficient bandpass filter may be fixed, but a mechanism for making the center frequency of the complex coefficient bandpass filter variable and controlling the frequency of the complex coefficient notch filter (for example, the normal angular frequency described above) The center frequency of the complex-coefficient bandpass filter may be controlled using an arithmetic processing unit that calculates a change value Ω _d ′.

Further, the first complex coefficient filter unit 722 of the first embodiment may be a complex coefficient band pass filter with a variable center frequency. In this case, it is necessary to provide a configuration for changing the center frequency f ₀ of the complex coefficient bandpass filter in accordance with the change in the system frequency f _s of the three-phase power system B.

In the second embodiment, the center frequency f ₀ of the complex coefficient bandpass filter of the first complex coefficient filter unit 722 is fixed. However, the center frequency f ₀ of the complex coefficient bandpass filter is also variable in the second embodiment. The center frequency f ₀ of the complex coefficient bandpass filter may be changed according to the change in the frequency f _s of the three-phase power system B, and the complex coefficient notch filter of the third embodiment is added to the complex coefficient bandpass filter. A combined configuration may be used. In these cases, the center frequency of the complex coefficient band-pass filter of the first complex coefficient filter unit 722 and the stop frequency of the complex coefficient notch filter are utilized using the configuration for controlling the center frequency f ₀ of the second complex coefficient filter unit 724 ′. Should be controlled. Of course, a complex coefficient notch filter may be combined with a complex coefficient bandpass filter with a fixed center frequency.

In the above embodiment, the QC mode frequency shift type isolated operation detection device has been described, but the present invention can also be applied to other frequency shift type isolated operation detection devices such as a slip mode frequency shift method. For example, in the slip mode frequency shift method, the frequency change rate calculation unit 71A of FIG. 1 is removed, and the reactive power fluctuation value generation unit 71B uses the frequency detected by the frequency detector 72 to change the reactive power target value Q _o . The value ΔQ may be set. Further, in the present invention using a frequency detection device that performs frequency detection using a complex coefficient filter for the frequency detector 72, if the frequency detection is performed and the isolated operation detection is performed using the fluctuation amount of the frequency, The present invention can be widely applied to other methods other than the shift method.

  In the above description, the power system B is a three-phase power system. However, it is needless to say that the present invention can be applied to a case where the power system B is a single-phase power system. In this case, what is necessary is just to make it the structure except the three-phase / two-phase conversion part 721 in the block diagram shown in FIG.3, FIG.15, FIG.17.

For single-phase, since the single-phase AC voltage v there is only one, is inputted to the first complex coefficient filter unit 722 as a real part u _r sampling data of the input data of the single-phase AC voltage v [k], “0” is input to the imaginary part u _j [k] of the input data. In addition, in the frequency detectors 72 and 72 ′ shown in FIG. 3, FIG. 15, and FIG. 17, the three-phase / two-phase converter 721 is removed, and the sampling data of the AC voltage v of any one of U, V, and W phases. May be input to the first complex coefficient filter unit 722 as the real part u _r [k] of the input data, and “0” may be input to the imaginary part u _j [k] of the input data.

In the first complex coefficient filter unit 722 using the complex coefficient filter, even when a single-phase AC voltage is input, the two-phase AC voltages v _r and v _j (sinusoidal wave and The first complex coefficient filter units 722 and 722 ′, the normalization unit 723, and the frequency extraction units 72B, 72B ′, and 72B ″ are output as shown in FIG. 3, FIG. 15, and FIG. This can be realized with the same configuration as the phase frequency detectors 72, 72 ′, 72B ″.

A Grid-connected inverter device B Three-phase power system 1 DC power supply 2 Inverter 3 Filter 4 Transformer 5 Circuit breaker 6 Control device 6a Bus voltage target value generation unit 6b Reactive power target value generation unit 6c Reactive power calculation unit 6d uvw-dq Conversion unit 6e, 6f, 6g, 6h PI compensation unit 6i, 6j Decoupling unit 6k dq-uvw conversion unit 6m PWM signal generation unit 7 Isolated operation detection device 71 Reactive power fluctuation controller 71A Frequency change rate calculation unit 71B Reactive power Fluctuation value generation unit 72, 72 ', 72 "Frequency detector 72A Disturbance removal unit 721 Three-phase / two-phase conversion unit 722, 722' First complex coefficient filter unit 723 Normalization processing unit 72B, 72B 'Frequency extraction unit 724 724 ′ second complex coefficient filter section 725 phase difference calculation section 726 frequency calculation section 73 islanding detector 8 AC voltage detector

Claims (9)

Frequency detecting means for detecting the frequency of the AC signal output from at least the grid-connected inverter device linked to the power system, and the grid-connected inverter device is brought into the single operation state based on the detection value of the frequency detecting means. In the isolated operation detection device of the grid interconnection inverter device, comprising an isolated operation detection means for detecting the transition,
The frequency detection means includes
AC signal detecting means for detecting the AC signal;
A first filter comprising a complex coefficient filter that removes the negative frequency component and the harmonic component of the fundamental wave contained in the AC signal detected by the AC signal detection means and outputs only the positive frequency component of the fundamental wave Means,
Frequency calculating means for calculating the frequency of the fundamental wave using the positive frequency component of the fundamental wave output from the first filter means;
An isolated operation detecting device for a grid-connected inverter device, comprising:   2. The isolated operation detection device for a grid-connected inverter device according to claim 1, wherein the complex coefficient filter of the first filter means is a complex coefficient band-pass filter whose center frequency is set to the frequency of the AC signal. The frequency calculation means includes
Second filter means comprising a multi-coefficient bandpass filter having a phase characteristic in which the center frequency is set to the frequency of the AC signal and the phase difference linearly changes in the passband;
Using the positive frequency component of the fundamental wave input to the second filter means and the positive frequency component of the fundamental wave output from the second filter means, the second filter according to a predetermined arithmetic expression A phase difference calculating means for calculating a phase difference between positive frequency components of the fundamental wave generated by the means;
The frequency of the positive frequency component of the fundamental wave input to the second filter means using a predetermined relational expression for obtaining the frequency from the phase difference in the pass band based on the phase characteristic and the phase difference. Frequency calculating means for calculating
The isolated operation detection apparatus of the grid connection inverter apparatus of Claim 1 or 2 containing this. The frequency calculation means includes
Second filter means comprising a multi-coefficient bandpass filter having a phase characteristic in which the center frequency is set to the frequency of the AC signal and the phase difference linearly changes in the passband;
Using the positive frequency component of the fundamental wave input to the second filter means and the positive frequency component of the fundamental wave output from the second filter means, the second filter according to a predetermined arithmetic expression A phase difference calculating means for calculating a phase difference between positive frequency components of the fundamental wave generated by the means;
A predetermined table indicating a relationship between a phase difference and a frequency in the passband set in advance based on the phase characteristics;
Frequency calculating means for calculating the frequency of the positive frequency component of the fundamental wave input to the second filter means by reading out the frequency corresponding to the phase difference calculated by the phase difference calculating means from the table; ,
The isolated operation detection apparatus of the grid connection inverter apparatus of Claim 1 or 2 containing this. The frequency calculation means includes
A passband type multi-coefficient filter having a phase characteristic in which the phase difference is zero at the center frequency, becomes negative in a frequency region higher than the center frequency, and becomes positive in a smaller frequency region, and the center frequency can be changed. A second filter means comprising:
Using the positive frequency component of the fundamental wave input to the second filter means and the positive frequency component of the fundamental wave output from the second filter means, the second filter according to a predetermined arithmetic expression A phase difference calculating means for calculating a phase difference between positive frequency components of the fundamental wave generated by the means;
If the phase difference calculated by the phase difference calculating means is positive, the center frequency of the second filter means is decreased by a predetermined change amount until the phase difference becomes zero, and the phase difference calculating means calculates the phase difference. If the phase difference is negative, the center frequency of the second filter means is increased by the amount of change until the phase difference becomes zero, and the center frequency after the change is output as a frequency detection value. Control means;
The isolated operation detection apparatus of the grid connection inverter apparatus of Claim 1 or 2 containing this.   The complex coefficient filter of the first filter means is based on the center frequency output from the center frequency control means, and has a stop frequency so that a negative frequency component and a harmonic component of a predetermined order with respect to the center frequency. The isolated operation detection device for a grid-connected inverter device according to claim 5, wherein the multi-coefficient notch filter is controlled. The grid-connected inverter device has a power major loop that controls the amount of reactive power output from the grid-connected inverter device,
Reactive power fluctuation value generating means for generating a reactive power fluctuation value for swinging the reactive power amount based on the frequency of the AC signal detected by the frequency detection means and feeding back to the power major loop, further comprising:
The islanding detection means calculates a fluctuation amount of the frequency detected by the frequency detection means, and detects that the grid-connected inverter device has shifted to an islanding state when the fluctuation amount exceeds a predetermined threshold. An isolated operation detection device for a grid-connected inverter device according to any one of claims 1 to 6.   The reactive power fluctuation value generation means calculates a frequency change rate using the frequency of the AC signal detected by the frequency detection means, and generates the reactive power fluctuation value that varies in proportion to the frequency change rate. An isolated operation detection device for a grid-connected inverter device according to claim 7.   The isolated operation detection device for a grid-connected inverter device according to any one of claims 1 to 8, wherein the AC signal is a single-phase or three-phase AC signal.