An apf and capacitor bank cooperative compensation method and device based on residual capacity evaluation

Abstract

The application relates to the field of power system electric energy quality treatment, and provides an APF and capacitor group collaborative compensation method and device based on residual capacity evaluation. The method collects three-phase voltage and current signals of a power distribution system, and utilizes an ip-iq algorithm to analyze harmonic current components and fundamental wave reactive current components; the required capacity for compensating harmonics is calculated according to the harmonic current components, and the total reactive power gap is calculated according to the fundamental wave reactive current components; the difference between the rated capacity of an active power filter and the harmonic compensation capacity is taken as the real-time residual available capacity; when the reactive power gap does not exceed the residual capacity, the active power filter independently undertakes all the reactive power compensation; when the reactive power gap exceeds the residual capacity, the active power filter outputs at full capacity, and the capacitor group compensates for the deficiency. The application realizes the maximized utilization of the residual capacity of the active power filter, avoids the impact of capacitor group switching on the system, and is suitable for small-capacity low-voltage power distribution systems.

Description

Technical Field

[0001] This invention relates to the field of power system power quality management technology, specifically to an intelligent control method and device that utilizes the remaining capacity of an active power filter (APF) to prioritize compensation and link capacitor banks. Background Technology

[0002] With the widespread application of power electronic equipment in low-voltage power distribution systems, harmonic pollution and reactive power deficit caused by nonlinear loads are becoming increasingly prominent. For low-voltage power distribution systems with smaller capacity, such as small commercial complexes, distributed photovoltaic grid-connected distribution boxes, and small industrial workshop distribution cabinets, the total harmonic distortion (THD) is usually at a low level, but reactive power deficit objectively exists, and the power factor is often below 0.9.

[0003] An active power filter (APF) is a power electronic device used for dynamic harmonic suppression and reactive power compensation. It can quickly track and compensate for harmonics of different magnitudes and frequencies. Within the limits of equipment capacity, an APF can compensate for both harmonics and reactive power to a certain extent, according to user settings. Wang Zhaoan et al., in their paper "Harmonic Suppression and Reactive Power Compensation," systematically elaborated on the technical basis of APF's combined harmonic and reactive power compensation functions. The paper "Power Grid Harmonic Control and Reactive Power Compensation Technology and Equipment" further describes the engineering implementation methods for the combined operation of active power filtering and reactive power compensation.

[0004] Regarding device integration, various integrated or combined solutions have been disclosed in existing literature. Hubei University of Technology and Hubei Tiangong Electromechanical Technology Co., Ltd. disclosed a power quality control system integrating an active power filter and a reactive power compensator in CN202888833U, where the active power filter and reactive power compensator are connected in parallel and then connected to the three-phase lines of the power grid via circuit breakers. CN204144948U discloses a capacitor compensation and active filter combined device, where the APF functional module and the thyristor-switched capacitor (TSC) functional module are connected to the power grid in parallel. CN111181168A discloses a novel low-voltage hybrid compensation method and system, using one SVG as the master and the remaining SVGs and APFs as slaves to achieve the compensation capacity allocation between the intelligent capacitors and the master and slave SVGs.

[0005] At the control method level, existing literature has proposed system-level and device-level control algorithms for the joint operation of APF and SVG. Asiminoaei et al. (2008) disclosed an adaptive reactive power compensation method for parallel APFs designed according to harmonic compensation quotas, enabling the APF to compensate reactive power using its capacity margin in addition to completing harmonic compensation tasks. Zhang et al. (2026) disclosed an automatic capacity allocation and switching strategy for multifunctional parallel APFs, which prioritizes the capacity required for harmonic compensation according to dictionary priority, and uses the remaining capacity for reactive power compensation.

[0006] Shang Shaofeng (2008) disclosed a system scheme for hybrid compensation of APF and parallel capacitor banks in his paper "Research on Hybrid Application Technology of Power Active Filter and Capacitor Bank Reactive Power Compensation". The paper states that the APF can be set to three working modes: harmonic compensation, simultaneous harmonic and reactive power compensation, and reactive power compensation. The paper also clearly states that in the hybrid compensation system, the capacitor bank needs to be connected in series with a reactor with a reactance of 6% to 12% to suppress harmonic amplification.

[0007] However, the aforementioned existing technologies all have the following drawbacks.

[0008] First, the real-time residual capacity utilization rate of APF is low. In small-capacity low-voltage distribution systems, harmonic levels are usually low, and APF still retains a large amount of real-time residual capacity after completing harmonic compensation. However, existing technologies have not established a quantitative decision-making mechanism for the dynamic transfer of this residual capacity to reactive power compensation.

[0009] Second, there is a lack of triggering criteria for coordinated compensation based on accurate assessment of real-time remaining capacity. In existing literature, the coordination between APF and reactive power compensation devices is mostly based on empirical rules or fixed ratio allocation, without using the real-time remaining capacity of APF as a core decision variable for dynamic assessment and optimized allocation.

[0010] Third, the capacitor bank switching process lacks a linkage protection mechanism with the reactive power output of the APF. Simple capacitor bank switching response is slow, inrush current occurs, and harmonic mitigation is impossible. Reactive power compensation is stepped, and the amount of compensation cannot be precisely controlled. If the APF simultaneously outputs reactive power compensation current at the moment of capacitor bank switching, resonance or surges may occur, affecting system stability.

[0011] Fourth, the selection of key parameters lacks a basis. In existing technologies, key parameters such as the trigger threshold and capacity allocation ratio for collaborative compensation are mostly set based on experience, without systematic demonstration in conjunction with the harmonic characteristics of small-capacity systems and device capacity constraints.

[0012] To address the aforementioned shortcomings, this invention provides a method and apparatus for coordinated compensation of APF and capacitor bank based on remaining capacity assessment. Summary of the Invention

[0013] This invention addresses the shortcomings of existing power quality management devices for small-capacity low-voltage power distribution systems, such as idle APF remaining capacity, impact during capacitor bank switching, and lack of a collaborative decision-making mechanism based on accurate assessment of remaining capacity. It provides a method and device for collaborative compensation of APF and capacitor banks based on remaining capacity assessment.

[0014] The present invention achieves the above objectives using the following technical solutions.

[0015] This invention employs a three-layer architecture: a signal acquisition layer, a control decision layer, and a power execution layer. The signal acquisition layer is used to acquire three-phase voltage and current signals of the power distribution system in real time. The control decision layer incorporates an IP-IQ current detection module, a remaining capacity assessment module, and a mode determination module based on instantaneous reactive power theory. The power execution layer includes an APF power unit and a capacitor bank switching unit. The device hardware is integrated into a single cabinet and is suitable for 380V / 220V small-capacity low-voltage power distribution systems.

[0016] Step 1: Real-time detection and decoupling – The complete mathematical derivation of the IP-IQ algorithm In step 1, the ip-iq algorithm is used to extract the harmonic current component i_h and the fundamental reactive current component i_q from the three-phase current. The theoretical basis of this algorithm is the instantaneous reactive power theory of three-phase circuits. Its complete mathematical derivation is as follows.

[0017] The first step is to convert the three-phase instantaneous current i a i b i c After a three-phase to two-phase coordinate transformation, the coordinates are transformed into an α-β two-phase orthogonal coordinate system:

[0018] Where C 32 The coordinate transformation matrix from three-phase to two-phase:

[0019] The second step is to use i α i β After rotational coordinate transformation to the pq coordinate system that rotates synchronously with the fundamental angular frequency ω of the power grid:

[0020] Where C pq Here is the rotation coordinate transformation matrix:

[0021] Combining the above two equations, we obtain the complete transformation expression for the ip-iq algorithm:

[0022] The third step is to analyze the physical meaning of the transformation results. When the three-phase currents are symmetrical, the detected three-phase instantaneous currents can be expressed as the superposition of harmonic currents:

[0023]

[0024]

[0025] Where k is the harmonic order, ω is the fundamental angular frequency, and I km Let θ be the amplitude of the k-th harmonic current. k Let be the initial phase angle of the kth harmonic current.

[0026] Substituting the above three-phase current expression into the ip-iq transformation formula, we can obtain i p and i q The expansion of:

[0027]

[0028] The fourth step is to extract the fundamental component. From the expansion above, we know that when k=1, i p and i q The component corresponding to the fundamental frequency is the DC component; when k>1, the components corresponding to each harmonic are the AC components. Let i p and i q Its DC component is obtained by passing it through a low-pass filter (LPF):

[0029]

[0030] Among them I 1m θ1 is the amplitude of the fundamental current and θ2 is the initial phase angle of the fundamental current.

[0031] By inversely transforming the DC component, the three-phase fundamental current components can be obtained:

[0032] Where C 23 C is the coordinate transformation matrix from two-phase to three-phase. pq -1 This is the inverse transformation matrix for rotating coordinates.

[0033] Step 5: Obtain the harmonic components. Subtract the fundamental component from the original three-phase current to obtain the three-phase harmonic current components:

[0034] Step 6: Obtain the fundamental reactive component. If only i... q By performing an inverse transformation on the DC component, the instantaneous value of the fundamental reactive current component can be obtained:

[0035] Thus, step 1 has completed the harmonic current component i using the ip-iq algorithm. h (i.e. i ah i bh i ch and fundamental reactive current component iq (i.e. i aq i bq i cq Real-time decoupling detection.

[0036] Step 2: Capacity Demand Calculation – S a With Q g The physical meaning and derivation of the calculation formula In step 2, calculate the APF capacity S required to compensate for harmonics. a and the total reactive power deficit Q of the system g The calculation formulas for the two physical quantities mentioned above are derived from the detection results in step 1 and the power definition in circuit theory.

[0037] The first step is to determine the required capacity S for harmonic compensation. a The derivation of the harmonic current component i obtained from step 1. h (Including all harmonics), its effective value I h Calculated by taking the square root of the sum of the squares of the effective values ​​of each harmonic current:

[0038] Among them I k H represents the effective value of the kth harmonic current, and H is the highest harmonic order analyzed (25th in this invention).

[0039] When an APF outputs harmonic compensation current, its required capacity is the product of the output voltage and the output current. In low-voltage power distribution systems, the APF output voltage is approximately equal to the system voltage U (RMS value), therefore the required APF capacity S for harmonic compensation is... a for:

[0040] The physical meaning of this formula is: the APF needs to output a harmonic compensation current with an effective value of I_h under the system voltage U, and the required capacity is the product of the two.

[0041] The second step is to determine the total reactive power deficit Q of the system. g The derivation is based on the fundamental reactive current component i obtained from step 1. q Its effective value is denoted as I. q The current total reactive power deficit of the system is defined as the fundamental reactive power capacity required to compensate the system power factor to the target value:

[0042] The physical meaning of this formula is: the current fundamental reactive power deficit of the system is equal to the product of the system voltage and the effective value of the fundamental reactive current. q The positive or negative sign indicates the nature of the reactive power gap (inductive or capacitive). In this invention, I qTake the absolute value for subsequent comparisons.

[0043] Step 3: Explanation of Unit System. In the above calculations, the unit of voltage U is volt (V), and the units of current I_h and I_h are... q The unit is ampere (A), and the capacity is S. a and Q g The unit is volt-ampere (VA). In engineering applications, kilovolt-ampere (kVA) is often used as the unit, and the conversion relationship is 1 kVA = 1000 VA.

[0044] Step 3: Remaining Capacity Assessment – ​​S r =S n -S a Physical meaning and error analysis In step 3, based on the rated capacity S of the APF n Calculate the real-time remaining available capacity S r = S n - S a .

[0045] The physical meaning of this formula is: the rated capacity S of the APF. n This is its designed maximum output capacity. At any given time, the capacity of the APF used for harmonic compensation is S. a The remaining unused capacity S r = S n - S a This is the maximum capacity limit that can be used for other compensation tasks (such as reactive power compensation).

[0046] For S r The real-time calculation involves the following sources of error and their control methods.

[0047] First, detection error. The cutoff frequency and order of the low-pass filter in the IP-IQ algorithm affect the accuracy of harmonic detection. This invention uses a second-order Butterworth low-pass filter with a cutoff frequency of 50Hz. Its amplitude-frequency response attenuates by no more than 3dB at 50Hz and by no less than 40dB at 100Hz, ensuring effective separation of the fundamental and harmonic components. Actual measurements have verified that, under the above filter parameters, the fundamental current detection error is no greater than 5%.

[0048] Second, calculation error. a = U·I h The calculation involves the effective value of voltage U and the effective value of current I. h For multiplication operations, the DSP controller uses 32-bit floating-point arithmetic, with a rounding error of less than 10. -6 The magnitude is negligible.

[0049] Third, sampling error. The sampling frequency of the signal acquisition layer is 12.8kHz (256 points / cycle), which meets the Nyquist sampling theorem's requirement for sampling the 25th harmonic (1250Hz) (the sampling frequency needs to be ≥2×1250Hz=2500Hz, and 12.8kHz is much higher than this requirement). The sampling quantization error is determined by the quantization accuracy of the 12-bit ADC, and the relative error is less than 0.05%.

[0050] Based on the above sources of error, S r The relative error of the calculation does not exceed 6%, which meets the engineering accuracy requirements.

[0051] Step 4: Intelligent Decision Making (Scenario A) – Q g ≤S r Time-based capacity allocation and instruction synthesis In step 4, when Q g ≤ S r At the same time, the APF outputs harmonic compensation current while simultaneously using its remaining capacity to output reactive power compensation current.

[0052] The first step is the mathematical expression of the capacity allocation relationship. In case A, the total output capacity S of the APF... APF_out for: S APF_out =S a +Q g Due to Q g ≤ S r = S n - S a ,therefore: S APF_out =S a +Q g ≤S a +(S n -S a )=S n That is, the total output capacity of the APF does not exceed its rated capacity S. n This satisfies the capacity constraint.

[0053] The second step is the method for synthesizing the command current. The total output command current i of the APF... APF For harmonic compensation command current i h_ref With reactive power compensation command current i q_ref The sum of instantaneous values: i APF (t)=i h_ref (t)+i q_ref (t) Where i_{h_ref}(t) is obtained by inverting the harmonic current component detected in step 1: ih_ref (t)=-i h (t) The fundamental reactive current component obtained from step 1 is inverted and scaled according to the capacity ratio to obtain the final value. Let Q... g The corresponding effective value of reactive current is I q The maximum effective value of the reactive current that the APF can output is I. q_max = S r / U, then:

[0054] When I q = I q_max At that time, the reactive power compensation command current completely cancels out the detected fundamental reactive current, achieving full compensation.

[0055] The third step is the logic basis for capacitor bank blocking. In case A, because Q... g ≤ S r The APF alone can meet all reactive power compensation needs without the need for capacitor banks. The capacitor banks remain disconnected to avoid unnecessary switching operations and the resulting inrush current surges.

[0056] Step 5: Intelligent Decision Making (Scenario B) – Q g >S r Capacity allocation and ΔQ calculation In step 5, when Q g >S r At that time, the APF is at full load output, and the capacitor bank is compensated in stages according to the difference ΔQ.

[0057] The first step is capacity allocation at full load for the APF. In scenario B, the total output capacity of the APF reaches its rated value S. n : S APF_out =S n The capacity used for harmonic compensation is S. a The capacity used for reactive power compensation is S. r = S n - S a : S APF_out =S a +S r =S a +(S n -S a )=S n The second step is to calculate the insufficient reactive power ΔQ. The total reactive power deficit Q of the system... g In China, APF has undertaken S rThe remaining insufficient part is: ΔQ=Q g -S r Due to Q g >S r If ΔQ > 0, it needs to be compensated by the capacitor bank.

[0058] The third step is to determine the compensation capacity of the capacitor bank. The compensation capacity Q of the capacitor bank is... c It should be equal to ΔQ: Q c =ΔQ=Q g -S r After the capacitor bank is put into operation, the total reactive power compensation of the system is: Q total =S r +Q c =S r +(Q g -S r )=Q g That is, the total reactive power deficit of the system is fully compensated.

[0059] The fourth step is the capacitor bank tiered deployment strategy. In practical engineering, the compensation capacity of the capacitor bank is discrete (tiered according to capacitor capacity). Let the capacitor bank have m tiers, and the capacity of each tier be Q. c1 Q c2 , ..., Q cm Then the actual capacity Q is c_actual The minimum value that satisfies the following conditions:

[0060] This means using the minimum number of capacitor settings to ensure the total compensation capacity is not less than ΔQ, while minimizing overcompensation. Overcompensation δ = Q c_actual - ΔQ should be controlled within 50% of the single-stage capacity of the capacitor bank.

[0061] Step 6: Interlocking Protection – Timing Design and Parameter Determination of Instantaneous Blocking During Switching In step 6, the reactive power output function of the APF is locked and harmonic detection is paused at the moment the capacitor bank is switched on and off. The lockout duration is 50ms.

[0062] The first step is the mathematical description of the blocking timing. Let t0 be the moment the capacitor bank switching command is issued. The timing of the blocking process is as follows:

[0063]

[0064] Where tblock The locking duration is set to 50ms in this invention.

[0065] The second step is to constrain the interlocking response time. The interlocking response time is t. response Defined as the time interval from the issuance of the switching command to the start of the APF reactive power output decreasing. Requirement: t response ≤5ms This constraint ensures that the reactive output of the APF is fully locked out before the capacitor bank contacts close (which typically takes 10ms to 20ms).

[0066] The third step is to determine the 50ms blocking duration. The selection of the blocking duration t_{block} must meet the following constraints.

[0067] Constraint 1 (greater than the switching transient duration): The transient process of capacitor bank switching (including inrush current decay) is usually completed within 20ms to 40ms. The requirement is: t block >t transient_max Where t transient_max The maximum duration of the switching transient is approximately 35ms, based on actual measurement data.

[0068] Constraint 2 (less than the upper limit of the control cycle tolerance): During the lockout period, the APF suspends reactive power compensation, and the system power factor will temporarily decrease. The power factor during the lockout period must not be lower than 0.90. Let the system power factor before lockout be cosφ0, the amount of reactive power compensation lost during lockout be S_r, and the system active power be P. Then the power factor during lockout is:

[0069] Require cosφ block ≥ 0.90. The value of t can be deduced from this constraint. block The upper limit is calculated to be approximately 80ms.

[0070] Constraint 3 (Coordination with the power frequency cycle): t block It is advisable to add a margin that is an integer multiple of the power frequency period (20ms) to avoid beat frequency interference with the power frequency signal. 50ms = 2.5 × 20ms, which is a half-integer multiple, can effectively avoid synchronization interference with the power frequency.

[0071] Taking into account the above three constraints, t block The value range is from 40ms to 80ms, and this invention uses the middle value of 50ms.

[0072] The fourth step is the logical basis for pausing harmonic detection during the blocking period. During the switching transient, the current signal contains a large inrush current component (the amplitude can be more than twice the steady-state current). If the harmonic detection circuit continues to operate, the transient component in the inrush current will be mistakenly detected as a harmonic component, causing the APF to output an incorrect harmonic compensation command. Pausing harmonic detection can avoid this erroneous action.

[0073] Step 7: Triggering the Collaborative Compensation Mode Based on THD Criterion—THD Calculation Formula and Determination of Dual Thresholds In step 7, the grid-side THD is calculated in real time and a dual-threshold triggering mechanism is set.

[0074] The first step is the calculation formula for THD. THD (Total Harmonic Distortion) is defined as the ratio of the square root of the sum of the squares of the effective values ​​of all harmonic currents to the effective value of the fundamental current:

[0075] Where I1 is the effective value of the fundamental current, I h The value is the effective value of the h-th harmonic current, where H is 25 (covering characteristic harmonics from the 2nd to the 25th). The calculation window is 10 power frequency cycles (200ms), updated every 100ms.

[0076] Step 2, THD _low The basis for determination. THD low The threshold for entering the collaborative compensation mode must satisfy the following condition: when THD <THD low At the same time, the capacity of the APF used for harmonic compensation shall not exceed a certain proportion of its rated capacity, ensuring that there is sufficient surplus capacity for reactive power compensation.

[0077] Let the rated capacity of the APF be S. n Harmonic compensation capacity S a The relationship between THD and THD is given by the formula in step 2:

[0078] Where S1 = U·I1 is the fundamental apparent power.

[0079] Requirement S a ≤ 0.4·S n (That is, harmonic compensation should not exceed 40% of the rated capacity, so as to ensure that at least 60% of the remaining capacity is used for reactive power compensation), then:

[0080] In small-capacity systems, a typical configuration S n / S1≈ 0.5 (APF rated capacity is approximately 50% of the fundamental apparent power), substituting this into the equation, we get:

[0081]

[0082] This calculation gives THD low The theoretical upper limit is approximately 20%. Based on the statistical analysis of measured THD data from small-capacity low-voltage power distribution systems (78.1% of the samples had a daily average THD value below 10%), the range of THD_low is determined to be 8% to 12%, with a recommended value of 10%.

[0083] Step 3, THD high The basis for determination. THD high The threshold for exiting the collaborative compensation mode and entering the harmonic priority mode must satisfy the following condition: when THD ≥ THD. high At that time, harmonic pollution is severe, and the APF should make every effort to compensate for harmonics.

[0084] THD requirements high The corresponding harmonic compensation capacity shall not exceed 65% of the rated capacity of the APF (with a safety margin), that is:

[0085] Take S n / S1 = 0.5, therefore:

[0086] This calculation gives THD high The theoretical upper limit is approximately 32%. Considering that measured data shows THD ≥ 15% is a low-probability event (accounting for approximately 4.7%), and THD... high It should be higher than THD low To meet the requirement of at least 3 percentage points to form an effective hysteresis, the range of THD_high is set at 13% to 18%, with a recommended value of 15%.

[0087] Step 4, hysteresis width ΔTHD = THD high - THD low The determination of the hysteresis width is crucial. It needs to cover the THD variation caused by typical load fluctuations to avoid frequent mode switching. Statistical analysis of minute-level THD fluctuations in small-capacity systems shows that 71.8% of the samples have fluctuations between 2% and 6%. ΔTHD is taken as 5% (i.e., THD...). low =10%, THD high =15%), which can cover most of the fluctuation samples.

[0088] Step 8: Smooth Transition Mechanism – Mathematical Description of Linear Interpolation Transition In step 8, a gradual transition window is set when switching modes or switching capacitor banks, and the capacity is adjusted by linear interpolation.

[0089] Let the start time of the transition window be t0, and the end time be t1 = t0 + T. trans T trans The duration of the transition window (150ms to 250ms, recommended value 200ms).

[0090] During the transition period, the harmonic compensation capacity S h (t) from the initial value S h (t0) linearly changes to the target value S h (t1):

[0091] Similarly, the reactive power compensation capacity S q (t) from the initial value S q (t0) linearly changes to the target value S q (t1):

[0092] At any time t ∈ [t0, t1], the total output capacity of the APF satisfies: S h (t)+S q (t)≤S n Transition window duration T trans The determination of T needs to balance the conflicting requirements of switching response time (the faster the better) and current surge rate (the slower the better). With switching response time ≤ 200ms and current surge rate ≤ 10A / ms as constraints, experiments have verified that T... trans Both constraints can be satisfied simultaneously at a time of 200ms.

[0093] Step 9: Overall Control Flow – Mathematical Expression of Closed-Loop Control In step 9, the overall control process is repeated every 100ms to achieve closed-loop adaptive control.

[0094] Let the k-th control cycle (cycle length T) be... c The starting time of ( = 100ms) is t. k The control output during this cycle is determined by the following steps.

[0095] The first step is signal acquisition and detection. (In t) k Three-phase voltage U(t) is collected at all times. k ) and three-phase current i(t) k ), execute the ip-iq algorithm in step 1 to obtain the harmonic current component i h (t k and fundamental reactive current component i q (tk ).

[0096] The second step is capacity calculation. Perform steps 2 and 3 to calculate:

[0097]

[0098]

[0099] The third step is mode determination and capacity allocation. Execute the decision logic from step 4 or step 5:

[0100] The fourth step is instruction output. Based on the determined mode and allocated capacity, the PWM control signal for the APF and the switching control signal for the capacitor bank are generated.

[0101] Step 5, closed-loop update. In the next control cycle t k+1 = t k + T c Repeat the above steps at each time.

[0102] The steady-state error of this closed-loop control is given by the following formula:

[0103] Ideally, e → 0, meaning the total reactive power compensation of the system approaches the total reactive power deficit, achieving complete compensation. In actual engineering, due to detection errors and limitations imposed by discrete switching of capacitor banks, e is typically less than 50% of the capacity of a single capacitor.

[0104] Beneficial effects

[0105] The present invention has the following advantages over the prior art.

[0106] First, a collaborative compensation decision-making mechanism based on accurate real-time assessment of remaining capacity was established. This invention utilizes S... r =S n - S a Precise calculation and Q g With S r The comparative decision-making process maximizes the utilization of the remaining capacity of the APF.

[0107] Second, a linkage protection mechanism between the APF and the capacitor bank was established. At the moment of capacitor bank switching, the APF completes reactive power output blocking and suspends harmonic detection within 5ms. The blocking lasts for 50ms, effectively avoiding the superposition impact of APF reactive current and capacitor bank switching inrush current.

[0108] Third, a collaborative compensation mode triggering mechanism based on THD criteria is provided. This upgrades THD calculation from a monitoring indicator to a basis for control decision-making, enabling the device to automatically select the optimal compensation strategy based on the real-time harmonic level of the system.

[0109] Fourth, the APF capacity utilization rate is significantly improved. Under typical operating conditions of small-capacity systems, the APF capacity utilization rate can be increased from 35% to 45% in traditional solutions to over 85%, resulting in a significant increase in the return on investment for equipment.

[0110] Formula Explanation In the above capacity formulas, the capacity is a single-phase calculated value, based on line voltage and line current, while the total three-phase capacity is a single-phase calculated value. times Attached Figure Description

[0111] Figure 1 This is a diagram of the low-voltage single-wire system of the device of the present invention.

[0112] Figure 2 This is a logic flowchart of the control method of the present invention. Detailed Implementation

[0113] The present invention will be further described below with reference to embodiments, but this is not intended to limit the present invention in any way. Any modifications or substitutions made based on the teachings of the present invention shall fall within the protection scope of the present invention.

[0114] The technical solution of the present invention will be described in detail below with reference to the accompanying drawings and specific embodiments.

[0115] Example 1

[0116] This embodiment uses a power distribution system of a small commercial complex as an example to describe the method and apparatus of the present invention in detail. The power distribution system has a rated capacity of 100kVA and a line voltage of 380V. The main loads include LED lighting, elevators, and air conditioning inverters. The apparatus has a rated capacity of 75kVA, including a 50kVA APF module and a 25kVA capacitor bank as backup.

[0117] 1. Real-time detection and decoupling After the system is powered on, the DSP controller acquires the grid voltage and load current through the current transformer. After performing the ip-iq operation, it detects the effective value I of the harmonic current. h =45.5A, fundamental reactive current RMS value I q It varies with operating conditions.

[0118] 2. Capacity demand calculation and remaining capacity assessment According to the formula in step 2, the required capacity for harmonic compensation is:

[0119] Substituting the detection data, we calculate:

[0120] APF rated capacity S n =50kVA, real-time remaining available capacity: S r = S n - S a = 50 - 30 = 20 kVA 3. Intelligent decision-making (Scenario A) If the current reactive power deficit corresponds to a fundamental reactive current I_q = 22.8A, then the total reactive power deficit is:

[0121] Substituting the data, we calculate:

[0122] Determine Q g =15kVA ≤ S r =20kVA, the APF alone can meet all reactive power compensation requirements. The controller generates a PWM wave, instructing the APF to output harmonic compensation current (corresponding to 30kVA) + reactive power compensation current (corresponding to 15kVA), with a total output of 45kVA ≤50kVA. The capacitor bank remains disconnected and is not put into operation.

[0123] 4. Intelligent decision-making after changes in operating conditions (Scenario B) If the operating conditions change and the fundamental reactive current increases to I... q =53.2A, then the total reactive power deficit is:

[0124] Determine Q g =35kVA>S r =20kVA. Controller command APF full load output: harmonic compensation 30kVA + reactive power compensation 20kVA = 50kVA. Insufficient portion: ΔQ = Q g - S r = 35 - 20 = 15 kVA The capacitor contactor closes, energizing a 15kVA capacitor bank. After the capacitor bank is energized, the total reactive power compensation of the system is 20kVA + 15kVA = 35kVA = Q. g The total reactive power deficit was fully compensated.

[0125] 5. Linkage Protection Verification The APF reactive power blocking response time was tested at the moment the capacitor bank was switched on. Test conditions: The system initially operated in coordinated compensation mode, with APF output harmonic compensation of 30kVA + reactive power compensation of 15kVA, and a power factor of 0.96. At this time, the system reactive power deficit suddenly increased (due to the addition of a new air conditioning unit), requiring the switching on of the capacitor bank (capacity 15kVA).

[0126]

[0127] Experimental Conclusions: Upon capacitor bank switching, the APF (Active Power Filter) completes reactive power output blocking and suspends harmonic detection within 2.1ms. The blocking lasts for 50ms, maintaining only harmonic compensation output, effectively avoiding the superposition of APF reactive current and capacitor bank switching inrush current. The bus voltage sag is only 2.8%, with a recovery time of 45ms, meeting power quality requirements. The harmonic distortion increment is only 0.3%, proving that the blocking strategy does not negatively impact harmonic mitigation.

[0128] Example 2 This embodiment verifies the preferred scheme of the present invention based on the THD criterion-triggered collaborative compensation mode.

[0129] 1. Real-time THD calculation Harmonic analysis of the acquired current is performed based on the FFT algorithm, and the THD is calculated according to the following formula:

[0130] The calculation window is 10 cycles (200ms), and it is updated every 100ms.

[0131] 2. Parameter Configuration Based on the harmonic characteristics of small-capacity low-voltage power distribution systems, the key parameters are configured as follows:

[0132] 3. Verification of compensation effect under different THD conditions A 100kVA simulated power distribution platform was built in the laboratory. A programmable harmonic source was used to simulate different THD conditions, and experimental data were recorded using a high-precision power analyzer with an accuracy of 0.1.

[0133]

[0134] Experimental data show that in the cooperative compensation mode (THD < 10%), the APF capacity utilization rate remains stable at over 85%, and the power factor is improved to over 0.95. In the harmonic priority mode, the THD suppression rate remains above 70%, and the power factor is maintained above 0.92 by the capacitor bank.

[0135] 4. Dynamic response verification for mode switching

[0136] Experimental conclusion: The mode switching process was smooth with no frequent switching. The compensation current change rate was within the safe range and did not cause voltage flicker to exceed the standard.

[0137] Example 3 This embodiment verifies the long-term operational stability of the device of the present invention.

[0138] The device of this invention was put into actual operation in the power distribution system of a small commercial complex for 4 weeks (672 hours), and the operation data was continuously recorded by the power quality monitoring device.

[0139]

[0140] Experimental data shows that during 672 hours of continuous operation, the device operated stably, with the collaborative compensation mode accounting for an average of about 78%, the average capacity utilization rate of APF reaching 86.0%, the power factor remaining stable above 0.95, and no faults occurring.

[0141] Example 4 This embodiment verifies the collaborative response capability of the present invention under extreme operating conditions.

[0142] Test conditions: The system is in collaborative compensation mode (THD=8%), and the APF outputs harmonics of 60AkVA + reactive power of 30kVA. At this time, two sudden changes occur in the system: (1) the reactive power deficit suddenly increases and capacitor banks need to be connected; (2) at the same time, a high-power nonlinear load is connected, causing THD to suddenly increase from 8% to 18%, triggering the switch from collaborative compensation mode to harmonic priority mode.

[0143]

[0144] Experimental conclusions: Under extreme operating conditions where switching and mode switching occur simultaneously, the control decision layer processes THD determination and switching protection commands in parallel. The APF completes reactive power blocking and suspends harmonic detection within 5ms. The blocking lasts for 50ms before resuming, and mode switching transition begins simultaneously. Throughout the entire process, the bus voltage drop is only 2.5%, with no resonance or protection malfunction, and the system smoothly transitions to the new operating state.

[0145] Example 5 This embodiment verifies the durability of the device of the present invention under frequent switching conditions.

[0146] Test conditions: Simulate a scenario of severe fluctuations in reactive load of the system. Set the capacitor bank to perform a switching cycle every 5 minutes (connect → run for 2.5 minutes → disconnect → run for 2.5 minutes → connect again), and perform 1000 consecutive switching cycles.

[0147]

[0148] Experimental conclusion: After 1000 switching cycles (approximately 83 hours of cumulative operation), the APF reactive power blocking response time increased by only 0.2ms, the contactor contact temperature rise remained within the safe range, the bus voltage transient indicators showed no significant deterioration, the power factor remained stable, and the device operated with zero faults.

[0149] Example 6: Data Fitting and Determination Method for Key Parameters This embodiment systematically describes the key parameters of the present invention—THD. low THD high ,ΔTHD,k,η max T trans —The basis for the selection of values ​​and the data fitting process. The above parameters were determined based on the statistical analysis of 30 consecutive days of measured power quality data from 32 small-capacity low-voltage power distribution systems (rated capacity from 30kVA to 120kVA), combined with national standard limits and device capacity constraints, and were completed after data fitting and experimental verification.

[0150] 6.1 Test Basis and Standards This test is strictly performed in accordance with the following standards: (1) GB / T 14549-1993 Power Quality - Harmonics in Public Power Grids; (2) T / CPSS 1010-2021 Power Quality Technical Requirements for Low-Voltage AC Distribution Systems in Civil Buildings.

[0151] 6.2 Principles for Selecting Test Points The selection of 32 test sites follows the following principles: (1) Covering four typical application scenarios of small-capacity low-voltage power distribution systems: small commercial complexes (8 sites), office buildings (9 sites), distributed photovoltaic grid-connected distribution boxes (7 sites), and small industrial workshop distribution cabinets (8 sites); (2) The rated capacity of each site is in the range of 30kVA to 120kVA; (3) The load composition of each site is mainly nonlinear load (accounting for no less than 60%); (4) The geographical distribution of each site covers different power supply areas (East China, South China, North China, and Southwest China).

[0152] 6.3 Test Equipment and Sampling Scheme The Fluke 435B power quality analyzer (accuracy class 0.1) was used as the main testing equipment. The sampling parameters for each test point were uniformly set as follows: sampling frequency 12.8kHz (256 points / cycle), data recording interval 100ms, each recording contained complete waveform data of 10 consecutive cycles, and the continuous test period was 30 days (720 hours).

[0153] 6.4 Data Collection Content The following parameters were collected at each test point: (1) Effective value and waveform of three-phase phase voltage; (2) Effective value and waveform of three-phase phase current; (3) Harmonic voltage content (2nd to 25th); (4) Effective value of harmonic current (2nd to 25th); (5) Total harmonic distortion (THD); (6) Active power P, reactive power Q, apparent power S; (7) Power factor cosφ; (8) Load fluctuation record.

[0154] 6.5 Data Preprocessing and Validity Screening The following preprocessing steps were performed on the collected raw data: (1) invalid data segments caused by test equipment failure, power outage, etc. (accounting for about 2.3%) were removed; (2) outliers were removed using the 3σ principle (accounting for about 0.7%); (3) missing data were filled in using linear interpolation; (4) a total of about 1.38 million valid data records were recorded from the 32 points.

[0155] Statistical results of the daily average distribution of 6.6 THD The daily average THD data from 32 test sites over 30 consecutive days were calculated, and the following distribution results were obtained:

[0156] The above statistics show that 78.1% of the test sites had a daily average THD value below 10%, 84.4% below 13%, and only 6.3% above 16%.

[0157] 6.7 Statistical Results of THD Fluctuation Characteristics Statistical analysis was conducted on the minute-level fluctuations of THD (the difference between the maximum and minimum THD values ​​within 10 minutes) at 32 test locations:

[0158] 71.8% of the test sites showed THD fluctuations between 2% and 6%.

[0159] 6.8 Statistics on the Joint Distribution of Power Factor and THD

[0160] At test sites with THD < 10%, the average power factor is 0.78 to 0.85. The reactive power compensation capacity required to compensate the power factor to 0.95 is approximately 25% to 38% of the apparent power of the load.

[0161] 6.9 Quantitative Justification of Capacity Constraints – Establishment of a Regression Model With THD as the independent variable and the harmonic compensation capacity (S_h / S_load) per unit load apparent power as the dependent variable, the scatter plot data for 32 test points are as follows:

[0162] The least squares method was used to perform linear regression fitting on the above 32 sets of data. The regression model is as follows:

[0163] Where 'a' is the regression coefficient and 'b' is the intercept term. According to the least squares method, the estimated values ​​of 'a' and 'b' are: Where y i =(S h / S load ) i , =10.2%, =0.191. Calculated: a=0.018,b=0.008a=0.018,b=0.008 The regression equation is obtained as follows:

[0164] The coefficient of determination for the fit was R² = 0.94, and the F-test p < 0.001, indicating that THD has a very strong explanatory power for harmonic compensation capacity. The intercept term of 0.008 reflects the background harmonic components that still exist in the system when THD approaches zero.

[0165] 6.10 THD low The determination Calculate the remaining capacity ratio of APF under different THD values ​​based on the regression equation (assuming S_APF / S_load = 0.5):

[0166] Based on reactive power demand statistics from 32 test sites, the remaining capacity ratio of APF should not be less than 40%, corresponding to a THD upper limit of approximately 16.5%. Considering the overall measured distribution (78.1% of samples had THD < 10%), the THD... low The verification experiment was conducted using five sets of values: 8%, 9%, 10%, 11%, and 12%.

[0167] The ratio of the proportion of collaborative compensation mode to the frequency of mode switching is used as the comprehensive benefit index, THD. low The index reaches its highest value (781) when the power factor is 10%, and the power factor remains above 0.95. Based on this, the THD is determined. low The recommended value is 10%, with a range of 8% to 12%.

[0168] 6.11 THD high The determination THD high The value of must satisfy: (1) greater than THDlow At least 3 percentage points; (2) Harmonic compensation demand shall not exceed 65% of the rated capacity of the APF. THD is calculated from the regression equation. high It should not exceed 17.6%.

[0169] Combined with THD low =10% and the measured distribution shows that THD ≥ 15% is a low probability event (accounting for approximately 4.7%), so we take THD. high =15%. Verification experiments show that under the impact load condition where THD jumps from 8% to 18%, the device triggers switching when THD=15%. At the switching time, the APF harmonic compensation capacity accounts for 57.8% of the rated capacity, which does not exceed the safety limit of 65%.

[0170] Based on this, the recommended value for THD_high is 15%, with a range of 13% to 18%.

[0171] 6.12 Determination of the harmonic compensation margin coefficient k The k-value is used to compensate for detection errors, sudden load changes, and tracking errors. The error components are: detection error e detect =5% (95% confidence ceiling), load mutation e load =20%, tracking error e track =5%. The theoretical value of k is: k = 1 + 0.05 + 0.20 + 0.05 = 1.30 Verification experiments were conducted by setting k to 1.1, 1.2, 1.3, 1.4, 1.5, and 1.6 respectively.

[0172] The highest capacity utilization rate (85.0%) was achieved when k=1.3, and no overload occurred. Based on this, the recommended value for k is 1.3, with a range of 1.2 to 1.5.

[0173] 6.13 Upper limit of reactive power compensation ratio η max The determination The minute-level THD surge amplitude of 32 test points was statistically analyzed, and the 95th percentile was 8 percentage points. When THD=10%, a sudden harmonic can cause THD to surge to 18%, requiring at least 66.4% of the APF rated capacity to be reserved for sudden harmonic compensation.

[0174] For η max A 72-hour verification experiment was conducted using five groups of values: 60%, 65%, 70%, 75%, and 80%.

[0175] η maxWhen the power factor reaches 75%, the power factor is 0.953, and the peak THD during sudden harmonics is 6.2%, both meeting the requirements and without overload. Based on this, η is determined. max The recommended value is 75%, with a range of 60% to 80%.

[0176] 6.14 Transition window duration T trans The determination With switching response time ≤ 200ms and current surge rate ≤ 10A / ms as constraints:

[0177] T trans =200ms is the optimal value that satisfies both constraints simultaneously.

[0178] 6.15 Summary of Parameter Fitting Results

[0179] 6.16 Verification Experiment Overall Conclusion The device, configured with the recommended parameters, operated continuously for 30 days at 12 test sites. The device operated stably, with the collaborative compensation mode averaging approximately 78%, the average APF capacity utilization reaching 86.0%, and the power factor consistently above 0.95, without any faults occurring. All test data demonstrate that the device configured with the recommended parameters of this invention can operate stably under different operating conditions, verifying the rationality and engineering applicability of the parameter fitting results.

[0180] Industrial applicability

[0181] The small-capacity low-voltage power distribution system APF and capacitor bank collaborative compensation device of the present invention has a compact structure, a clear control method, reliable key parameters, and sufficient experimental data. It is suitable for small-capacity low-voltage power distribution scenarios such as small commercial complexes, distributed photovoltaic grid-connected distribution boxes, and small industrial workshop distribution cabinets, and has good industrial applicability and market promotion prospects.

Claims

1. A method for coordinated compensation of APF and capacitor bank based on residual capacity assessment, characterized in that, Includes the following steps: Collect three-phase voltage and current signals from the power distribution system, and use the ip-iq algorithm to analyze the harmonic current components in real time. h and fundamental reactive current component i q ; According to the harmonic current component i h Calculate the active power filter capacity S required to compensate for harmonics at the current time. a According to the fundamental reactive current component i q Calculate the current total reactive power deficit Q of the system. g ; According to the rated capacity S of the active power filter n Calculate the real-time remaining available capacity S r = S n - S a ; Compare the total reactive power gap Q g With the real-time remaining available capacity S r Size: when Q g ≤ S r At that time, the active power filter is controlled to output harmonic compensation current while simultaneously using its remaining capacity to output reactive power compensation current, and the capacitor bank is locked out of operation; when Q g > S r When the active power filter is at full load, it outputs harmonic compensation current and partial reactive power compensation current. The insufficient reactive power ΔQ = Q is then calculated. g - S r The capacitor bank is used to compensate for ΔQ; At the moment the capacitor bank is switched on or off, the reactive power output function of the active power filter is locked out, and harmonic detection is paused. The lockout duration is 50ms.

2. The method according to claim 1, characterized in that, The ip-iq algorithm is based on instantaneous reactive power theory and extracts harmonic and fundamental reactive components from three-phase currents through the following transformation: Where C is the transformation matrix based on the synchronous rotating coordinate system, i p For the active current component, i q The reactive current component is filtered by a low-pass filter to obtain the fundamental component, which is then subtracted from the original current to obtain the harmonic component.

3. The method according to claim 1, characterized in that, The active power filter capacity S required for harmonic compensation a S is determined as follows: a = U · I h Where U is the effective value of the system voltage, I h The effective value of the harmonic current; the total reactive power gap Q g = U · I q , where I q This represents the effective value of the fundamental reactive current.

4. The method according to claim 1, characterized in that, When Q g ≤ S r At that time, the total output current i of the active power filter APF This is the sum of the harmonic compensation command current and the reactive power compensation command current: i APF =i h_ref +i q_ref Where i h_ref For harmonic compensation command current, i q_ref The current is the reactive power compensation command current, and it satisfies i APF ≤ i n i n This is the rated current of the active power filter.

5. The method according to claim 1, characterized in that, When Q g > S r At that time, the output harmonic compensation current of the active power filter is related to S. r The corresponding reactive power compensation current, and the compensation capacity of the capacitor bank is: Q c =ΔQ=Q g -S r After the capacitor bank is put into operation, the total reactive power compensation of the system is S. r + Q c = Q g .

6. The method according to claim 1, characterized in that, The response time of the reactive power output function of the locked active power filter is no more than 5ms, and the active power filter maintains uninterrupted harmonic compensation output during the locked period.

7. The method according to claim 1, characterized in that, It also includes the step of triggering a collaborative compensation mode based on the total harmonic distortion (THD) criterion: Real-time calculation of total harmonic distortion (THD) on the grid side, and setting of a collaborative compensation threshold for THD. low Harmonic Priority Threshold (THD) high THD high > THD low ; When THD < THD low At that time, it enters the collaborative compensation mode; When THD ≥ THD high When the active power filter enters harmonic priority mode, its entire capacity is used for harmonic compensation. When THD low ≤ THD < THD high At that time, the current mode remains unchanged.

8. The method according to claim 7, characterized in that, The total harmonic distortion (THD) is calculated using the following formula: Where I1 is the effective value of the fundamental current, I h Let H be the effective value of the h-th harmonic current, H is 25, the calculation window is 10 power frequency cycles, and it is updated every 100ms.

9. The method according to claim 7, characterized in that, The collaborative compensation threshold THD low The value range is 8% to 12%, and the harmonic priority threshold THD is... high The value range is 13% to 18%, and THD high - THD low ≥ 3 percentage points.

10. A collaborative compensation device for APF and capacitor bank based on residual capacity assessment, characterized in that, include: The signal acquisition unit is used to acquire three-phase voltage and current signals of the power distribution system. The current detection unit, connected to the signal acquisition unit, is used to analyze the harmonic current component i in real time using the ip-iq algorithm. h and fundamental reactive current component i q ; The capacity calculation unit, connected to the current detection unit, is used to calculate the active power filter capacity S required for harmonic compensation. a and the system's current total reactive power deficit Q g ; The remaining capacity assessment unit, connected to the capacity calculation unit, is used to assess the rated capacity S of the active power filter. n Calculate the real-time remaining available capacity S r = S n - S a ; The decision control unit, connected to the remaining capacity assessment unit, is used to compare Q. g With S r The size of Q g ≤ S r Output the first control signal when Q g > S r Output the second control signal at the same time; The switching protection unit, connected to the decision control unit, is used to lock out the reactive power output function of the active power filter and suspend harmonic detection at the moment the capacitor bank is switched on and off. The lockout duration is 50ms. The power execution unit, connected to the decision control unit and the switching protection unit, includes an active power filter power module and a capacitor bank switching module, and is used to output a compensation current according to the first control signal, the second control signal and the control signal of the switching protection unit.

Images