Island detection method with system oscillation locking and self-adaptive blind area identification

Abstract

The invention discloses a photovoltaic island detection method based on system oscillation locking and self-adaptive blind area elimination, and by developing a photovoltaic island detection method based on a harmonic voltage starting criterion and self-adaptive blind area elimination, the anti-island protection capability of a power distribution network is enhanced, and the reliability of a power distribution network relay protection device is improved. An active island detection method is improved by introducing a positive feedback coefficient, and self-adaptive improvement of a traditional injection frequency shift method is completed. The island effect blind area detection of the power distribution network accessed by the distributed photovoltaic system can be refined, the detection blind area is eliminated to the greatest extent, and the non-island working condition is effectively filtered. The positive feedback active injection current frequency shift method organically integrates sudden change detection, offset detection and blind area identification, so that anti-islanding protection has rapidness, reliability and practicability. The introduction of a self-adaptive injection criterion constructed based on the voltage amplitude frequency offset and a positive feedback coefficient can avoid the unnecessary start of the active injection current and reduce the influence of the active injection current on the electric energy quality.

Description

Technical Field

[0001] The present invention relates to the field of circuit devices, and in particular to a photovoltaic island detection method based on system oscillation blocking and adaptive blind zone elimination. Background Art

[0002] Under the background of the "dual carbon" strategy, the penetration rate of photovoltaic power in the distribution network has been increasing year by year. Designing an island detection scheme that can quickly identify the island state and maintain high power quality is an urgent problem to be solved for the safe and stable operation of the new power system. Regarding the island detection problem of photovoltaic access to the distribution network system, the protection of the grid connection point often faces two contradictory choices: if the one-sided pursuit of reliability will lead to unnecessary large-scale disconnection of photovoltaic power; if the one-sided pursuit of safety will sacrifice the speed of island detection, posing new challenges to the safe and stable operation of the power grid. The latest specification for the speed of island protection, "Technical Specifications for Grid Connection of Distributed Power Sources" (GB / T 33593-2017), requires that the distribution network system connected to distributed power sources (including photovoltaics) can quickly and accurately detect the island operation state, and when the island occurs, the time to cut off the distributed power source should not exceed 2s in principle.

[0003] At present, island detection methods are mainly divided into three categories: remote method, passive method and active method. Among them, the remote method has technical advantages such as no blind area, little impact on power quality, and good speed, but it is also accompanied by high installation, operation and maintenance costs, and it is difficult to promote in actual engineering applications. The passive method has a simple principle and low configuration cost, but it is difficult to calculate a reasonable setting value in a complex distribution network system, and it is also accompanied by a large detection blind area. The active method extracts new fault features by actively injecting distorted current disturbances. The detection blind area is small and has high reliability, but it will sacrifice a certain amount of power quality.

[0004] Therefore, there is an urgent need for a hybrid islanding detection solution that can effectively avoid unnecessary startup of the active method, reduce the impact of system oscillation on islanding detection, and adaptively identify the blind area of ​​islanding detection and timely activate the active injection current frequency shifting method. Summary of the invention

[0005] This invention takes the island detection model of distributed photovoltaic large-scale access to the distribution network as the research object, and intends to enhance the anti-islanding protection capability of the distribution network and improve the reliability of the distribution network relay protection device by developing a photovoltaic island detection method based on harmonic voltage startup criteria and adaptive blind zone elimination. By introducing a positive feedback coefficient to improve the active island detection method, the adaptive improvement of the traditional injection frequency shift method is completed. The technical solution is as follows:

[0006] A photovoltaic island detection method based on system oscillation blocking and adaptive blind zone elimination uses the following detection steps, including:

[0007] S1: Perform power blind spot detection of the photovoltaic system grid connection point to calculate the grid blind spot identification result in real time; at the same time, perform total harmonic distortion detection of the grid connection point voltage to detect whether the total harmonic distortion mutation of the grid connection point voltage exceeds the set limit. If the total harmonic distortion mutation of the voltage exceeds the set limit, proceed to step S2;

[0008] S2: If the voltage total harmonic distortion mutation exceeds the set limit, stop blind spot identification, and perform a second detection on the grid connection point voltage total harmonic distortion mutation after a delay to detect whether the voltage total harmonic distortion mutation at the grid connection point exceeds the set limit again. If the voltage total harmonic distortion mutation exceeds the set limit for the second time, proceed to step S3;

[0009] S3: Perform voltage frequency offset blind zone detection at the grid connection point of the photovoltaic system to detect whether the active power and reactive power injected at the grid connection point are in the detection blind zone. If the injected power is in the detection blind zone, proceed to step S4; if the injected power is outside the detection blind zone, proceed to step S5;

[0010] S4: using an active current injection frequency shift method with a positive feedback coefficient to detect, so as to eliminate the blind spot of the grid connection point detection. If there is no blind spot after detection, proceed to step S5;

[0011] S5: Perform voltage amplitude and frequency deviation detection on the grid-connected point of the photovoltaic system to detect whether the voltage amplitude and frequency on the grid-connected point exceed the set limit within the prescribed time. If the set limit is exceeded, the anti-islanding protection device is activated to cut off the connection between the photovoltaic system and the local load.

[0012] Furthermore, the voltage total harmonic distortion does not exceed twice the previous sampling point. If it exceeds, the voltage total harmonic distortion of the previous sampling point is recorded as the reference value; then the detection is delayed for 100ms and again detected. If the detected voltage total harmonic distortion still exceeds twice the reference value, the system oscillation lockout is closed.

[0013] Furthermore, in the case of system oscillation caused by load changes, the active injection current frequency shifting method in step S4 is not started, thereby avoiding power quality loss caused by unnecessary circumstances.

[0014] Furthermore, the set limit value of the voltage amplitude offset detection is 0.85 times to 1.1 times the rated voltage.

[0015] Furthermore, the offset limit of the frequency offset detection is 49.5 Hz to 50.5 Hz.

[0016] Furthermore, the chopping factor in the active current injection frequency shifting method is defined so that the initial phase shift direction of the inverter active current injection is consistent with the phase shift direction caused by the local load of the system.

[0017] Furthermore, the positive feedback coefficient is introduced into the active injection current frequency shifting method. f0 ×C norm The feedforward coefficient k is obtained by the coordinate system method, and the chopping factor in the frequency shift method is corrected.

[0018] A central data processor, characterized in that the central data processor is used to run programs and control logic signals, wherein when the program is running, any one of the photovoltaic island detection methods based on system oscillation locking and adaptive blind zone elimination is executed.

[0019] A microcomputer-readable storage medium, characterized in that the microcomputer-readable storage medium includes a stored program, and when the program is running, the device where the microcomputer-readable storage medium is located is controlled to execute any one of the photovoltaic island detection methods based on system oscillation locking and adaptive blind zone elimination.

[0020] Compared with the prior art, the present invention has the following significant advantages:

[0021] 1. It can improve the blind spot detection of the islanding effect of the distribution network to which the distributed photovoltaic system is connected, eliminate the detection blind spots to the greatest extent, and effectively filter out non-islanding conditions.

[0022] 2. The positive feedback active injection current frequency shift method is organically integrated with mutation detection, offset detection and blind spot identification, making the anti-islanding protection fast, reliable and practical.

[0023] 3. The introduction of adaptive injection criterion and positive feedback coefficient based on voltage amplitude frequency offset can avoid unnecessary start-up of active injection current and reduce its impact on power quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 It is a photovoltaic island detection model.

[0025] Figure 2 The present invention is a flow chart of a hybrid island detection method based on system oscillation blocking and adaptive blind spot identification.

[0026] Figure 3 This is a system oscillation blocking solution based on the harmonic voltage detection method of the present invention.

[0027] Figure 4 This is a detection blind area diagram based on voltage offset and frequency offset of the present invention.

[0028] FIG5( a ) is a schematic diagram of the first case of the active injection current frequency shifting method of the present invention.

[0029] FIG5( b ) is a schematic diagram of the second case of the active injection current frequency shifting method of the present invention. DETAILED DESCRIPTION

[0030] The present invention is further described in detail below in conjunction with the accompanying drawings and examples. The following description of the specific implementation method of the present invention is to make the purpose, technical solution and advantages of the present invention more clearly described, and does not limit all technical details of the present invention.

[0031] like Figure 1 This is a photovoltaic island detection model, and the detection method of the present invention is specifically described by taking the primary system as an example.

[0032] This detection method requires the installation of a voltage sensor and a grid synchronous vector measurement unit (PMU) at the photovoltaic access point PCC to measure and transmit the voltage amplitude U at the PCC in real time. pcc and phase θ pcc information.

[0033] This detection method requires the installation of a current sensor at the photovoltaic output point to measure the output current amplitude I of the photovoltaic in real time. pv Since a phase-locked loop is used between the photovoltaic control output and the voltage at the access point PCC, the current vector of the photovoltaic output and the voltage vector at the access point PCC are in phase, and there is no need to measure the current phase information of the photovoltaic output.

[0034] This detection method requires the installation of current sensors and grid synchronous vector measurement units (PMUs) at the load access points to measure and transmit the load current amplitude I at the load access points in real time. load and phase θ load information.

[0035] This detection method requires the installation of a central data processing module at the island protection installation site, which specifically includes a phase detector: used to calculate the load power angle information; a blind spot calculation module: used to calculate and form an island detection blind spot range database; a logic judgment module: used to calculate whether to perform active current injection frequency shifting method and the logic information of whether the island protection is activated.

[0036] The detection method uses a synchronization method with a unified clock. The measuring element at the grid access point PCC is synchronized once every 1s by the 1PPS signal issued by the central data processing module; the measuring element at the load access point is synchronized once every 1s by the 1PPS signal issued by the central data processing module. As a result, the absolute error value between the global sampling pulses does not exceed 2μs, achieving strict synchronization of global measurement.

[0037] The detection method obtains the frequency of the voltage at the photovoltaic access point PCC in real time through the time interval of the voltage zero-crossing signal received by the central data processing module. According to the voltage zero-crossing once every half cycle, the frequency information of the grid-connected point at the current moment can be obtained: pcc =12(t0-t1), where t0 is the moment when the current voltage crosses zero, and t1 is the moment when the voltage last crossed zero.

[0038] This detection method is a photovoltaic island detection method based on system oscillation locking and adaptive blind zone elimination, which includes a system oscillation locking method based on harmonic voltage detection with delay; an island blind zone detection method based on voltage and frequency offset; and an active injection current frequency shift detection method that introduces a positive feedback coefficient.

[0039] The following will describe the specific implementation steps of the detection method, explain the specific implementation methods and characteristics of different technologies in the detection method, and finally analyze the technical points and hardware requirements in each step.

[0040] First, if Figure 2 As shown in the flow chart, the system will use the delayed harmonic voltage detection technology to determine whether the system is experiencing system oscillation caused by load changes at this moment. If so, the system oscillation lockout is turned on and the active injection current frequency shift method is not allowed to start; if not, the system oscillation lockout is turned off and enters the blind spot judgment stage.

[0041] This method requires detecting the harmonic voltage at the photovoltaic access point PCC once every half cycle and delaying repeated detection to determine whether the system has experienced system oscillation at this time. The details are as follows: the system performs a harmonic detection on the system every half cycle (i.e. 0.01s) to obtain the amplitude of each harmonic voltage of the grid-connected point voltage (seventh and below), and the total harmonic distortion (THD) parameter is calculated based on this.

[0042] Furthermore, the voltage total harmonic distortion is compared with the previous sampling value. If the total harmonic distortion detected this time is twice or more than the total harmonic distortion detected last time, the previous detection time is recorded as t0, and delayed by 100ms, and the above detection process is repeated. If the total harmonic distortion obtained at time (t0+100) is still twice or more than the total harmonic distortion at time t0, it can be determined that the harmonic voltage mutation is not caused by system oscillation, and the system oscillation lockout is closed.

[0043] The system oscillation locking method based on the sudden change of voltage total harmonic distortion is characterized by: being able to avoid the risk of misjudgment of locking detection caused by the access of a large number of nonlinear loads to the power grid; at the same time, using the system oscillation locking method with delayed harmonic voltage detection, the sudden increase of voltage total harmonic distortion caused by system oscillation will be quickly reduced in a short time, therefore, the detected voltage total harmonic distortion will return to a normal level after a 100ms delay, ensuring that the protection can be reliably locked in the case of system oscillation, thereby avoiding false start-up of the active injection current frequency shifting method.

[0044] This method requires a harmonic voltage detection every half cycle, so a phase detector needs to be installed at the PCC to send a signal every time the voltage crosses zero, and then a harmonic voltage measurement sequence is performed; when the method detects that the total harmonic distortion of the voltage detected this time is 2 times or more of the total harmonic distortion of the voltage detected last time, it needs to record the current time t0 and apply it to the blind spot detection module, so a data storage device and a signal communication device are required. When the method detects that the current voltage total harmonic distortion exceeds the limit and delays 100ms to detect the limit again, it is judged that the harmonic voltage mutation at this time is not caused by system oscillation, and the system oscillation lock is turned off, and a logic signal k1 is transmitted. When the oscillation lock is started, k1=0; when the oscillation lock is turned off, k1=1, the logic signal at this time is stored and sent, so a data storage device and a signal communication device are required.

[0045] Secondly, if Figure 2 As shown in the flow chart, the blind zone calculation data collection is synchronized with the harmonic voltage detection collection, and the blind zone calculation is performed in the central data processing module.

[0046] Specifically, whenever the phase detector at the PCC detects a voltage zero-crossing signal, the following information is collected synchronously with the harmonic voltage detection: the voltage sensor at the photovoltaic access point PCC and the grid synchronous vector measurement unit (PMU) collect the voltage amplitude Upcc and phase θpcc information at the PCC and transmit them to the central data processing module through the signal communication device; similarly, the current sensor at the photovoltaic output point collects the photovoltaic output current amplitude I pv And transmit it to the central data processing module through the signal communication device; similarly, the current sensor at the load access point and the grid synchronous vector measurement unit (PMU) collect the load current amplitude I at the load access point load and phase θ load Similarly, the central data processing module calculates f according to the current time when the voltage zero-crossing signal is received and the time when the voltage zero-crossing signal was received last time. pcc =1 / 2(t0-t1).

[0047] Furthermore, the central data processing module calculates in real time the active power ΔP injected by the grid into the PCC, the reactive power ΔQ injected by the grid into the PCC, and the active power P injected by the photovoltaic into the PCC based on the above collected information. load , the reactive power Q injected by photovoltaic into PCC load , and store.

[0048] When the system is in normal operation, that is, the harmonic voltage detection device at the PCC does not detect that the voltage total harmonic distortion mutation exceeds the limit, the blind spot calculation data acquisition is synchronized with the harmonic voltage detection acquisition, that is, the data is collected synchronously every half cycle; the harmonic voltage detection device at the PCC detects that the voltage total harmonic distortion mutation exceeds the limit, and records the current time t0, and then delays 100ms. During this period, the blind spot calculation data also stops being sent to the central data processing module to avoid information stacking. Furthermore, if the harmonic voltage detected again after a delay of 100ms does not exceed the limit, it is judged that the sudden change of the total harmonic distortion of the voltage is caused by the system oscillation, the oscillation lockout is still started, k1=0 is set, and k1 is sent to the central data processing module, the harmonic voltage detection and collection is still restored to once every half cycle, and the blind spot calculation data collection is still synchronized; if the sudden change of the total harmonic distortion of the voltage detected again after a delay of 100ms still exceeds the limit, it is judged that the sudden change of the total harmonic distortion of the voltage is not caused by the system oscillation, the system oscillation lockout is closed, k1=1 is set, and k1 is sent to the central data processing module, and the recorded time t0 is also transmitted synchronously.

[0049] Furthermore, when the central data processing module receives the logic signal k1=0, it determines that the voltage total harmonic distortion mutation is caused by system oscillation. At this time, the system oscillation lock is turned on and the blind zone calculation module is not started; when the central data processing module receives the logic signal k1=1, it determines that the harmonic voltage mutation is not caused by system oscillation, and starts the blind zone calculation module. By traversing the database at the recorded measurement time t0, the following information ΔP, ΔQ, P calculated at time t0 is obtained. load , Q load , according to (U N U PCC.max ) 2 -1≤ΔPP DG ≤(U N U PCC.min ) 2 -1;Q f (f PCC.min f N -f N f PCC.min )≤(1+ΔPP DG ) -1 ΔQP DG ≤Q f (f PCC.max fN -f N f PCC.max ) and draw the detection blind area map as shown in Figure 4 When the measurement point at time t0 is outside the detection blind zone, it is determined that the system is not within the range of the island detection blind zone, k2=0 is set, and the island detection blind zone coefficient k2 is stored, and a pulse for updating the island blind zone coefficient k2 is sent to the logic processing module; when the measurement point at time t0 is within the detection blind zone (including the boundary), it is determined that the system is within the range of the island detection blind zone, k2=1 is set, and the island detection blind zone coefficient k2 is stored, and a pulse for updating the island blind zone coefficient k2 is sent to the logic processing module.

[0050] Specifically, the island detection range refers to the load range corresponding to the power system correctly identifying the island operation state. If the matching degree between photovoltaic and load is high, the voltage amplitude and frequency of the system will change less in the island operation state, which is represented by the following formula: and Therefore, a high degree of matching between PV output and load may lead to failure in the detection of voltage offset and frequency offset of the system. The load range corresponding to the inability to correctly identify the island operation state is defined as the detection blind zone (non-detection zone, NDZ).

[0051] Furthermore, the total impedance of the RLC parallel load is expressed as: Where R, L, and C are the load resistance, inductance, and capacitance values, respectively. When the capacitive impedance and inductive impedance of the load cancel each other out, the system becomes a resistive load and resonance occurs in the system. The angular frequency at this time is called the resonant angular frequency (expressed as w res ), we can get The value of NDZ corresponds to the load range that makes island detection invalid, that is, the larger the calculated NDZ value is, the larger the load range that makes island detection invalid is.

[0052] Specifically, to solve the coupling problem, Q f0 ×C norm The coordinate system method describes the coordinate system from a new perspective and defines Q f0 is the load quality factor, expressed as Qf0 = R / (w0L). Since the load quality factor defined by the load resistance and inductance is independent of the load capacitance, and the use of the per-unit value of the inductance as the vertical axis, the coordinate space is decoupled. The coordinate system method is used to express the island detection blind area more concisely and clearly. Without further explanation, the following description of the island detection blind area will be based on Q f0 ×C norm Coordinate system method.

[0053] Secondly, if Figure 2 As shown in the flow chart, when the central data processing module receives the pulse signal of the blind spot detection module coefficient k2 update, it enters the logic judgment module. The system oscillation lockout coefficient k1 and the blind spot detection module coefficient k2 are calculated through the "AND" logic module. If the logic output signal k1∩k2=1, it means that the system oscillation lockout is closed at this time, and the system is within the detection blind spot range. At this time, the improved active current frequency shift detection method is started; if the logic output signal k1∩k2=0, it means that the system oscillation lockout is closed at this time, and the system is not within the detection blind spot range. At this time, the island blind spot detection stage based on voltage and frequency offset is started.

[0054] Further, such as Figure 2 As shown in the flow chart, the improved active injection current frequency shift detection method is used for the system in the island detection blind area, and the distorted current is actively injected into the grid through the photovoltaic inverter, so that the voltage frequency at the photovoltaic access point changes. When the grid is operating normally, the grid frequency at the grid connection point is kept stable due to the clamping effect of the large grid; if the photovoltaic is in an island operation state at this time, the voltage level of the PCC is out of the control of the large grid, and will be affected by the active injection current, so that the frequency changes with the distorted current, so that the voltage frequency shift element can detect the island operation state at this time.

[0055] Specifically, as shown in the principle diagram of the active injection current frequency shifting method in FIG5 , t z is the zero output time in the photovoltaic output waveform through the inverter, and T is the voltage cycle of the photovoltaic access point. Therefore, the active injection current frequency shifting method will have two situations. When the current half-wave output by the inverter is ahead of the voltage half-wave, the current waveform will have a period of dead zone after the half-wave is completed until the voltage half-wave is also completed; when the current half-wave output by the inverter lags behind the voltage half-wave, the inverter control forces the current to be set to zero and then forces the next cycle to start. Figure 5(a) is the first situation, and Figure 5(b) is the second situation.

[0056] The chopping factor (ChoppingFactor, C f ), defined as: C f =2t z / T, from this definition we know that C f The larger the absolute value, the greater the disturbance injected by the active current frequency shift method. f When C is positive, the frequency shift method will increase the frequency of the injected current; when C f When it is a negative value, the frequency shift method will make the frequency of the injected current smaller.

[0057] From the above description, it can be seen that the actual operation and detection effect of the active injection current frequency shift method and the chopping factor C fThis method determines the chopping factor C from the following perspectives: f Selection:

[0058] First, reduce the range of the dead zone (NDZ). Since the active injection current frequency shifting method will cause the change of the coupling point voltage frequency, the positive feedback of the voltage frequency can effectively accelerate the offset. If the system is in the grid-connected state, due to the clamping effect of the large power grid frequency, the grid-connected point voltage frequency will not fluctuate after the injection of the distorted current. Therefore, the absolute value of the chopping coefficient will not increase continuously due to the iteration, but will oscillate around the initial chopping value. The grid-connected point voltage frequency in the island state is affected by the photovoltaic output distortion current. Therefore, the positive feedback iteration of the chopping factor can effectively save the time of island detection. The use of the chopping factor with voltage frequency positive feedback is related to the offset value of the grid-connected point voltage frequency and the rated voltage frequency: Generally, the larger the offset value, the larger the absolute value of the chopping factor. The specific relationship is as follows: C f =C f0 +k(f pcc -f grid ), where C f0 is the initial chopping factor, k is the positive feedback coefficient, f pcc is the voltage frequency at the coupling point, f grid The frequency of the grid during normal operation. The principle of using the active injection current frequency shift method (AFDPF) with voltage frequency positive feedback is to obtain the frequency information of the grid connection point once every half cycle, substitute it into the formula of the chopping factor for iteration, update the injected current waveform output by the inverter, and continue to measure the frequency information of the grid connection point in the next cycle to complete an iteration cycle. Repeat the above iteration process within the specified time range until the frequency exceeds the limit or the island detection limit time ends.

[0059] Secondly, the use of active current injection frequency shifting (AFDPF) with voltage frequency positive feedback can reduce the range of the dead zone (NDZ), but this method cannot completely eliminate the dead zone. The phase shift caused by the power factor angle of the load is opposite to the phase shift caused by the frequency shifting method. When the two effects cancel each other out, the output current I "seen" by the phase-locked loop (PLL) is pv With common point voltage U pcc In phase, the following formula should be established: tan -1 [R(wC-1wL)]=πC f 2. At this time, the current I actively injected by the inverter pv The frequency will no longer be adjusted. If the voltage frequency value at this time is still within the limit, the measurement dead zone problem will still occur. Therefore, the selection of the positive feedback coefficient should be calibrated within a certain range.

[0060] Specifically, this paper adopts Q f0 ×C norm The coordinate system method describes the dead zone range of the active current injection method. f0 is the load quality factor at rated frequency, which is equal to R / (w N L); C norm is the per-unit value of the load capacitance. When an island occurs, the grid-connected point voltage angular frequency w is expressed as w N +Δw, where Δw is the difference between the grid-connected point voltage angular frequency and the grid rated angular frequency. The load capacitance can be expressed as C = C norm C0 = C0 (1 + ΔC); It is called resonant capacitor. Substituting the parameters after the island occurs into the final island steady-state formula, we can get (w N +Δw)(1+ΔC)C0-1 / [(w N +Δw)L]=(1 / R)tan(πC f / 2), the system frequency range is set to 49.5Hz~50.5Hz, and it is substituted into the above formula to obtain the capacitance range of detection failure when the island occurs in steady state. Substituting it into the above formula, the following conclusion can be drawn: If C norm If the upper limit is smaller than the lower limit, the system frequency will not appear in the range of 49.5Hz to 50.5Hz in the island steady state of active current injection, and the steady-state failure problem will not occur. At this time, the value range of the positive feedback coefficient when there is no steady-state detection blind zone can be solved as follows: k>(2 / k3)*(0.08Q f0 / π-2C f0 ), generally, the selection of the positive feedback coefficient needs to comprehensively consider multiple factors such as the speed of island detection and power quality. The larger the value of k, the faster the corresponding harmonic detection speed. However, if the value of k is too large, the power quality of the power system will be affected due to the injection of too many harmonics. Therefore, it is necessary to control the total harmonic distortion (THD) of the system.

[0061] Third, selecting a sufficiently large positive feedback coefficient k can avoid the blind spot of steady-state island detection, but at the same time it will cause the total harmonic distortion (THD) of the system to increase. Therefore, the value range of the positive feedback coefficient k cannot be too large. The following is a quantitative analysis of the chopping coefficient and the total harmonic distortion (THD) caused to the system: Fourier expansion of the waveform of the inverter output voltage can produce a convergent Fourier series, in which the coefficients are expressed as: Among them, w f =2π / T, m represents the mth harmonic. Substituting the output current function of the inverter into the formula, the content of each harmonic contained in the current can be obtained. Furthermore, the calculation formula of the total harmonic content is calculated according to the following formula: The value of THD varies with C f The absolute value of increases, and the change trend can be approximated as a linear relationship. Generally, the total harmonic content THD in the output current should not exceed 5%. Initial chopping factor C f0 The selection of C cannot be too large. f0 The larger the value of k is, the wider the range of k will be, but it will also lead to an increase in the total harmonic content. Therefore, in actual engineering practice, the appropriate initial chopping coefficient C should be selected according to the total harmonic content value. f0 , thereby determining the positive feedback coefficient according to the load conditions.

[0062] Fourthly, the influence of the dilution effect of the power grid is considered. Considering that the initial phase angle of the inverter is related to the chopping factor, when the chopping direction of the system is opposite to the load direction, the voltage phase shift caused by the chopping and the phase shift caused by the load cancel each other out, so that the influence of the active injection current on the system voltage frequency is diluted, which may cause the island protection to fail. Therefore, the adaptive parameters are introduced to make the initial phase angle chopping direction and the phase shift direction caused by the load the same. The C of the improved method is used. f0 The positive and negative of the previous coefficients are consistent with the positive and negative of the previous coefficients of the positive feedback. It is used to ensure that the offset direction is the same as the frequency offset direction caused by the load in the case of an island, so as to reduce the impact of the dilution effect. The sign function can be used to f0 Define, that is, C f =C f0 sign(f pcc -f grid )+k*sign(f pcc -f grid ), where the sign function is the mathematical symbol function. When f pcc -f grid When ≥0, the output of the sign function is positive; f pcc -f grid When <0, the output result of the sign function is negative, thus ensuring that the initial chopping direction is the same as the positive feedback chopping direction.

[0063] Fifth, the frequency offset is introduced to correct the chopping coefficient. The specific implementation process is as follows: when the voltage frequency of the PCC point is less than 0.2Hz different from the voltage frequency of the normal operation of the power grid, a small disturbance is added; when the voltage frequency difference exceeds 0.2Hz, a larger disturbance is added. At this time, the values ​​of the chopping factor and the voltage positive feedback coefficient are increased to accelerate the voltage frequency change of the PCC point. The above method can improve the operating stability of the system in the grid-connected state, and can also improve the speed of disturbance injection and detection in the island operation state.

[0064] Specifically, a frequency deviation reflection coefficient k3 is added in front of the positive feedback coefficient. k3 is a piecewise step function, which is specifically defined as follows: k3 is the frequency deviation reflection coefficient. When the change amplitude of the grid-connected point voltage frequency is within ±0.2Hz (the normal fluctuation range of the grid frequency), k3 is set to 1. Otherwise, it shows a linear change trend with the frequency deviation. The specific value is determined by the following formula: Then the expression of the positive feedback part becomes C f =(C f +k3k)sign(f pcc -f grid ).

[0065] In summary, the present invention proposes a hybrid improved active injection current detection method with voltage positive feedback. The chopping coefficient adopted comprehensively considers system oscillation lockout, blind area identification, steady-state island blind area and frequency deviation feedback. The expression of the chopping factor obtained by comprehensive consideration is as follows: C f =k1k2(C f +k3k)sign(f pcc -f grid ), where k1 is the oscillation blocking coefficient. When the system oscillation blocking is on, k1=0, and when the system oscillation blocking is off, k1=1; k2 is the blind spot identification coefficient. When the system is not in the detection blind spot, k2=0, and when the system is in the detection blind spot, k2=1. k3 is the frequency deviation reflection coefficient. When the voltage frequency at the grid connection point changes within ±0.2Hz (the normal fluctuation range of the grid frequency), k3 is set to 1. Otherwise, it shows a linear change trend with the frequency deviation. The specific value is determined by the following formula: Substituting into the steady-state blind zone analysis, the value of the positive feedback coefficient can be obtained. In general, the load quality factor does not exceed 2.5, that is, when the load quality factor is 2.5, the value of Q f0 . At the same time, the chopping factor C f0 Select the appropriate chopping factor C for the total harmonic content (THD) corresponding to different values ​​of f0 And the positive feedback coefficient k. That is, the appropriate positive feedback coefficient k is selected under the premise of comprehensively considering the island steady-state blind area and ensuring that the power quality meets the requirements of the regulations.

[0066] As mentioned above, at the end of each voltage half cycle, a voltage frequency offset detection is performed in the central data processing center, which is synchronized with the iteration of the chopping coefficient, as follows: Get the real-time frequency value f of the coupling point voltage pcc ; Set the coupling point voltage frequency f pcc With rated voltage frequency f NThe difference is iterated into the chopping factor calculation formula for positive feedback process; the chopping factor is substituted into the voltage inverter, and disturbance is added to the grid-connected system based on this; it is determined whether the voltage frequency of the current coupling point exceeds the limit. If it exceeds the threshold, the system is in an island state; otherwise, return to the first step to continue iteration; if the number of iterations exceeds the set time, the coupling point voltage frequency is still within the limit, then the loop is automatically exited. Generally, in order to ensure the reliable operation of island protection while ensuring the power quality of the power grid, the prescribed detection limit practice is 2000ms. The process of detecting whether the voltage frequency of the coupling point exceeds the limit is the same as the specific process of not being in the island detection blind area, except that the iterative process of active current injection is not used in the non-island detection blind area. Therefore, the specific process is specifically introduced in the next process.

[0067] Further, such as Figure 2 As shown in the flow chart, the voltage frequency offset detection is used for the system that is not in the island detection blind area. As above, at the end of each voltage half cycle, the central data processing module first calculates the voltage frequency of the current grid connection point, that is, f pcc =12(t0-t1), enter the logic judgment module, which is an "AND" module composed of two size comparators, that is, both satisfy f pcc ≤50.5Hz and f pcc ≥49.5Hz, at this time the island protection action coefficient K p The output is 0, indicating that the island operation state is not detected at this time; otherwise, the island protection action coefficient K p The output is 1, indicating that the system has detected an island operation state.

[0068] Further, such as Figure 2 As shown in the flow chart, when the central data processing module receives the island protection action coefficient K p Updated pulse, make a judgment: if K p = 0, and the 200th pulse has not been received, indicating that no island has been detected at this time, and the specified island detection time has not yet arrived, the detection process continues, the system oscillation lock coefficient is still set to 1, the system oscillation lock is still closed, and the detection continues; if K p = 0, and the current is the 200th pulse received, indicating that no island is detected at this time, and the specified island detection time has arrived, the detection process is completed, the system oscillation lock coefficient is set to 0, the system oscillation lock is turned on, and the oscillation lock detection phase is restarted; if K p =1, indicating that island operation is detected within the limited detection time. At this time, an island operation signal should be sent to the island protection circuit breaker immediately to act on the circuit breaker to complete the tripping task.

[0069] In summary, an embodiment of the present invention provides a hybrid island detection method with system oscillation blocking and adaptive blind zone identification, including:

[0070] System oscillation blocking method based on delayed harmonic voltage detection; the voltage total harmonic distortion shall not exceed 2 times of the previous sampling point. If it exceeds, the voltage total harmonic distortion of the previous sampling point is recorded as the reference value. Then, the detection is repeated after a delay of 100ms. If the detected voltage total harmonic distortion still exceeds 2 times of the reference value, the system oscillation blocking is closed.

[0071] Furthermore, the risk of misjudgment of lockout detection caused by the access of a large number of nonlinear loads to the power grid can be avoided; at the same time, the system oscillation lockout method with delayed harmonic voltage detection is used. Since the sudden increase in total harmonic distortion of voltage caused by system oscillation will be quickly reduced in a short time, the total harmonic distortion of voltage detected after a 100ms delay will return to normal levels, ensuring that the protection can be reliably locked in the case of system oscillation, thereby avoiding false startup of the active injection current frequency shifting method.

[0072] Furthermore, in the case of system oscillation caused by load changes, the active injection current frequency shifting method is not activated, thereby avoiding power quality loss caused in unnecessary situations.

[0073] The island blind area detection method based on voltage and frequency offset, the voltage offset limit is 0.85 to 1.1 times the rated voltage, and the frequency offset limit is 49.5 Hz to 50.5 Hz; the voltage and current sensors and the grid synchronization measurement unit are connected to the grid connection point to detect the state parameters of the grid connection point in real time, and the load power angle parameters are calculated using the phase detector. The synchronization method of the unified clock is introduced to achieve strict synchronization of global measurements, and the sampling pulse error is within 2 μs;

[0074] The active injection current frequency shift detection method with positive feedback coefficient is introduced to make the initial phase shift direction of the inverter consistent with the shift direction caused by the local load of the system by improving the positive and negative properties of the initial chopping factor; f0 ×C norm The coordinate system method is used to obtain the feedforward coefficient k, and the chopping factor in the frequency shift method is corrected; the load quality factor does not exceed 2.5; compared with the traditional AFD detection method and the AFDPF detection method, the detection speed is greatly improved without a detection blind spot, and the power quality is guaranteed to meet the requirements of the regulations.

[0075] S1: Perform power blind spot detection at the photovoltaic system grid connection point to calculate the grid blind spot identification result in real time; at the same time, perform total harmonic distortion detection at the grid connection point voltage to detect whether the total harmonic distortion (THD) mutation amount of the grid connection point voltage exceeds the set limit value. If the total harmonic distortion mutation amount of the voltage exceeds the set limit value, enter step S2;

[0076] S2: If the voltage total harmonic distortion mutation exceeds the set limit, the blind spot identification is stopped, and the voltage total harmonic distortion mutation of the grid connection point is tested again after a delay to detect whether it exceeds the set limit again. If the voltage total harmonic distortion mutation exceeds the set limit for the second time, the process goes to step S3;

[0077] S3: Perform voltage frequency offset blind zone detection at the grid connection point of the photovoltaic system to detect whether the active power and reactive power injected at the grid connection point are in the detection blind zone. If the injected power is in the detection blind zone, proceed to step S4; if the injected power is outside the detection blind zone, proceed to step S5;

[0078] S4: Active injection current frequency shift detection with a positive feedback coefficient is used to eliminate the blind spot of the photovoltaic system grid connection point detection. If there is no blind spot after detection, proceed to step S5;

[0079] S5: Perform voltage amplitude and frequency deviation detection on the grid-connected point of the photovoltaic system to detect whether the voltage amplitude and frequency on the grid-connected point exceed the set limit within the prescribed time. If the set limit is exceeded, the anti-islanding protection device is activated to cut off the connection between the photovoltaic system and the local load.

[0080] Through the passive islanding detection method of S1, S2, S3, and S5 combined with the active injection current frequency shift detection method of S4 introducing a positive feedback coefficient, a comprehensive judgment is made on the islanding state to improve the accuracy of anti-islanding protection.

[0081] The present application also provides an electronic device, which may include: at least one processor and at least one memory.

[0082] The processor may include one or more central data processors, and the central data processor is used to run programs and control logic signals, wherein the program executes a photovoltaic island detection method based on system oscillation locking and adaptive blind zone elimination when running.

[0083] The processor uses various interfaces and lines to connect various parts within the entire server, and performs various functions of the server and processes data by running or executing instructions, programs, code sets or instruction sets stored in the memory, and calling data stored in the memory.

[0084] Optionally, the processor can be implemented in at least one hardware form of digital signal processing (DSP), field programmable gate array (FPGA), and programmable logic array (PLA). The processor can integrate one or a combination of a central processing unit (CPU), a graphics processing unit (GPU), and a modem. Among them, the CPU mainly processes the operating system, user interface, and application programs; the GPU is responsible for rendering and drawing the content to be displayed on the display; and the modem is used to process wireless communications. It is understandable that the above-mentioned modem may not be integrated into the processor, but may be implemented separately through a chip.

[0085] Among them, the memory may include a random access memory (Random Access Memory, RAM) or a read-only memory (Read-Only Memory). Optionally, the memory includes a non-transitory computer-readable storage medium. The memory can be used to store instructions, programs, codes, code sets or instruction sets. The memory may include a program storage area and a data storage area, wherein the program storage area may store instructions for implementing an operating system, instructions for at least one function (such as a touch function, a sound playback function, an image playback function, etc.), instructions for implementing the above-mentioned various method embodiments, etc.; the data storage area may store data involved in the above-mentioned various method embodiments, etc. The memory may also be optionally at least one storage device located away from the aforementioned processor. The memory as a computer storage medium may include an operating system, a network communication module, a user interface module, and an application program for a small incision thyroid surgery image acquisition system.

[0086] A microcomputer-readable storage medium includes a stored program. When the program is run, the device where the microcomputer-readable storage medium is located is controlled to execute a photovoltaic island detection method based on system oscillation locking and adaptive blind zone elimination.

[0087] In addition, each functional unit in the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0088] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable memory. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product, which is stored in a memory and includes several instructions for a computer device (which can be a personal computer, server or network device, etc.) to perform all or part of the steps of the various embodiments of the present application. The aforementioned memory includes: various media that can store program codes, such as USB flash drives, mobile hard drives, magnetic disks or optical disks.

[0089] The above is only an exemplary embodiment of the present disclosure, and the scope of the present disclosure cannot be limited thereto. That is, any equivalent changes and modifications made according to the teachings of the present disclosure are still within the scope of the present disclosure. After considering the disclosure of the specification and the truth of practice, it will be easy for those skilled in the art to think of other embodiments of the present disclosure. This application is intended to cover any modification, use or adaptation of the present disclosure, which follows the general principles of the present disclosure and includes common knowledge or customary technical means in the technical field that are not recorded in the present disclosure. The description and examples are regarded as exemplary only, and the scope and spirit of the present disclosure are defined by the claims.

Claims

1. A photovoltaic island detection method based on system oscillation blocking and adaptive blind zone elimination, characterized in that: Use the following testing steps, including: S1: Perform power blind spot detection of the photovoltaic system grid connection point to calculate the grid blind spot identification result in real time; at the same time, perform total harmonic distortion detection of the grid connection point voltage to detect whether the total harmonic distortion mutation of the grid connection point voltage exceeds the set limit. If the total harmonic distortion mutation of the voltage exceeds the set limit, proceed to step S2; S2: If the voltage total harmonic distortion mutation exceeds the set limit, stop blind spot identification, and perform a second detection on the grid connection point voltage total harmonic distortion mutation after a delay to detect whether the voltage total harmonic distortion mutation at the grid connection point exceeds the set limit again. If the voltage total harmonic distortion mutation exceeds the set limit for the second time, proceed to step S3; S3: Perform voltage frequency offset blind zone detection at the grid connection point of the photovoltaic system to detect whether the active power and reactive power injected at the grid connection point are in the detection blind zone. If the injected power is in the detection blind zone, proceed to step S4; if the injected power is outside the detection blind zone, proceed to step S5; S4: using an active current injection frequency shift method with a positive feedback coefficient to detect, so as to eliminate the blind spot of the grid connection point detection. If there is no blind spot after detection, proceed to step S5; S5: Perform voltage amplitude and frequency deviation detection on the grid-connected point of the photovoltaic system to detect whether the voltage amplitude and frequency on the grid-connected point exceed the set limit within the prescribed time. If the set limit is exceeded, the anti-islanding protection device is activated to cut off the connection between the photovoltaic system and the local load.

2. The photovoltaic island detection method based on system oscillation blocking and adaptive blind zone elimination according to claim 1 is characterized in that: The voltage total harmonic distortion is not more than twice of the previous sampling point. If it exceeds, the voltage total harmonic distortion of the previous sampling point is recorded as the reference value; then the detection is delayed for 100ms and tested again. If the detected voltage total harmonic distortion still exceeds twice of the reference value, the system oscillation lock is closed.

3. The photovoltaic island detection method based on system oscillation blocking and adaptive blind zone elimination according to claim 2 is characterized in that: In the case of system oscillation caused by load changes, the active injection current frequency shifting method in step S4 is not started, thereby avoiding power quality loss caused in unnecessary situations.

4. The photovoltaic island detection method based on system oscillation blocking and adaptive blind zone elimination according to claim 1 is characterized in that: The set limit value of the voltage amplitude deviation detection is 0.85 times to 1.1 times the rated voltage.

5. The photovoltaic island detection method based on system oscillation blocking and adaptive blind zone elimination according to claim 1 is characterized in that: The deviation limit of the frequency deviation detection is 49.5 Hz to 50.5 Hz.

6. The photovoltaic island detection method based on system oscillation blocking and adaptive blind zone elimination according to claim 1 is characterized in that: The chopping factor in the active current injection frequency shifting method is defined so that the initial phase shift direction of the active current injected by the inverter is consistent with the phase shift direction caused by the local load of the system.

7. The photovoltaic island detection method based on system oscillation blocking and adaptive blind zone elimination according to claim 1 is characterized in that: The positive feedback coefficient is introduced into the active injection current frequency shifting method, and the feedforward coefficient k is obtained according to the Qf0×Cnorm coordinate system method to correct the chopping factor in the frequency shifting method.

8. A central data processor, characterized in that: The central data processor is used to run programs and control logic signals, wherein when the program is run, any one of the photovoltaic island detection methods based on system oscillation locking and adaptive blind zone elimination is executed.

9. A microcomputer readable storage medium, characterized in that: The microcomputer-readable storage medium includes a stored program, and when the program is executed, the device where the microcomputer-readable storage medium is located is controlled to execute any one of the photovoltaic island detection methods based on system oscillation locking and adaptive blind zone elimination.

Images