An adaptive switching control method for low-voltage complete reactive power compensation device

Abstract

The application discloses a kind of low-voltage complete set reactive power compensation device's self-adapting switching control method, specifically relates to low-voltage distribution reactive compensation control technical field, the voltage, current, power factor, reactive power, harmonic parameter and the capacitor capacity that has been put into of acquisition low-voltage bus, based on power factor determines conventional compensation demand quantity;According to the short-time change characteristics of voltage, current and reactive power, load mutation precursor value is generated, and harmonic impedance close value is generated according to harmonic parameter and the equivalent impedance change after capacitor input;The switching prediction ability result is obtained by comprehensive evaluation of the two, and the state that the switching prediction ability is satisfied and the state that the switching prediction ability is insufficient are divided;When the switching prediction ability is satisfied, it is switched according to conventional compensation demand quantity, and when the switching prediction ability is insufficient, the compensation capacity, switching interval and capacitor branch are dynamically adjusted and then switched in stages;The application can reduce mis-switching, over-compensation, harmonic amplification and capacitor branch overcurrent risk, improve compensation stability.

Description

Technical Field

[0001] This invention relates to the field of low-voltage power distribution reactive power compensation control technology, specifically to an adaptive switching control method for a low-voltage complete reactive power compensation device. Background Technology

[0002] Low-voltage reactive power compensation devices are widely used in factories, commercial buildings, data centers, and low-voltage distribution substations. They primarily compensate for the inductive reactive power of low-voltage buses by switching capacitor banks, thereby improving the power factor, reducing line losses, and enhancing voltage quality. Existing devices typically collect parameters such as voltage, current, power factor, and reactive power of the low-voltage bus. When the power factor is lower than the set target or the reactive power exceeds the set range, the controller switches on the corresponding capacitor branches according to a preset sequence, cyclic mode, or capacity matching mode; when the power factor is advanced or the reactive power decreases, some capacitor branches are disconnected.

[0003] However, with the large-scale connection of frequency converters, rectifiers, welding equipment, charging equipment, photovoltaic inverters, and high-power motors with frequent start-stop cycles to the low-voltage distribution side, the load state of the low-voltage bus exhibits complex characteristics such as short-term abrupt changes, harmonic content fluctuations, and changes in equivalent impedance. In such scenarios, existing switching methods mostly rely on instantaneous power factor or instantaneous reactive power as the main judgment criteria. It is difficult to identify the precursors of impending load abrupt changes before the capacitor bank is connected, and it is also difficult to determine whether the capacitor branch to be connected will cause the bus equivalent impedance to approach the main harmonic frequency. Therefore, problems such as undercompensation misjudgment, excessive switching capacity, unreasonable switching intervals, or risky branches being connected may occur.

[0004] The above problems can lead to short-term power factor lead, bus voltage rise, harmonic voltage amplification, overcurrent heating in capacitor branches, and frequent switching of compensation cabinets after the capacitor bank is put into operation. In severe cases, it may also cause the upper-level protection to activate or the production load to stop. Summary of the Invention

[0005] The purpose of this invention is to provide an adaptive switching control method for a low-voltage complete reactive power compensation device to address the shortcomings in the prior art.

[0006] To achieve the above objectives, the present invention provides the following technical solution: an adaptive switching control method for a low-voltage complete reactive power compensation device, comprising: S1. Collect the voltage, current, power factor, reactive power, harmonic parameters and the capacity of the capacitors already in operation of the low-voltage bus, and determine the conventional compensation requirement based on the power factor. S2, generates load surge precursor values ​​based on the short-time variation characteristics of voltage, current and reactive power, and generates harmonic impedance approximation values ​​based on the harmonic parameters and the equivalent impedance change after capacitor connection. S3, comprehensively evaluate the load change precursor value and harmonic impedance proximity value to obtain the switching prediction capability result, and classify the prediction capability as satisfied or insufficient based on the result; S4. When the predicted capacity is met, the capacitor bank is switched on and off according to the conventional compensation requirement. When the predicted capacity is insufficient, the compensation capacity, switching interval and capacitor branch are dynamically adjusted according to the load change precursor value and harmonic impedance proximity value, and then switched in stages. After each stage of switching, the switching predicted capacity result is updated.

[0007] Preferably, determining the conventional compensation requirement based on the power factor includes: Using the in-phase zero-crossing moment of the low-voltage bus voltage as the sampling reference, the voltage and current are subjected to integer truncation, and the transient fluctuation range after the capacitor bank is put into or removed is eliminated to obtain steady-state sampling data. The fundamental voltage, fundamental current, and effective power factor used for compensation judgment are determined based on the steady-state sampling data. The effective power factor is compared with the preset target power factor, and the capacity deficit is determined by combining the reactive power and the capacity of the capacitors already in operation. The capacity deficit is then matched with the capacity of the capacitor branches to obtain the conventional compensation requirement.

[0008] Preferably, the transient fluctuation range after the capacitor bank is connected or disconnected includes: Record the connection or disconnection time of each capacitor branch, and simultaneously record the start and end cycles of the low-voltage bus current jump. Based on the cycle in which the input or cut-off time occurs, cycles containing current jumps and their adjacent transition cycles are eliminated. The phase-continuous integer-cycle data after removal are spliced ​​together in chronological order to serve as steady-state sampling data for determining the effective power factor and harmonic parameters.

[0009] Preferably, in S2, load surge precursor values ​​are generated based on the short-time variation characteristics of voltage, current, and reactive power, including: Using multiple consecutive in-phase zero-crossing cycles as short-time observation intervals, the effective voltage, effective current, and reactive power of the low-voltage bus are sorted cycle by cycle to obtain short-time voltage, short-time current, and short-time reactive power sequences. Based on the short-time voltage sequence, short-time current sequence, and short-time reactive power sequence, the voltage disturbance, current surge, and reactive power rise are determined respectively. The voltage disturbance, current surge, and reactive power rise are correlated in chronological order of occurrence, and the load surge precursor value is determined based on the correlation results and the number of continuous cycles.

[0010] Preferably, determining the load mutation precursor value based on the correlation results and the number of continuous cycles includes: When the frequency corresponding to the voltage disturbance is earlier than the frequency corresponding to the current surge, and the frequency corresponding to the current surge is earlier than the frequency corresponding to the reactive power rise, it is confirmed that there is a precursor to load energization. The number of continuous cycles of the load energizing precursor within a short observation interval is counted, and the number of times the voltage disturbance, current surge, and reactive power rise synchronously increase in adjacent cycles is determined. Based on the number of continuous cycles and the number of synchronous enhancements, the voltage disturbance, current surge, and reactive power rise are corrected to obtain the load change precursor value.

[0011] Preferably, in step S2, generating a harmonic impedance approximation value based on the harmonic parameters and the equivalent impedance change after capacitor connection includes: Based on the conventional compensation requirements, the combination of capacitor branches to be put into operation is determined, and the harmonic order, harmonic voltage and harmonic current with the highest proportion at the low voltage bus are extracted from the harmonic parameters. The equivalent impedance of the bus before switching is determined based on the harmonic voltage and harmonic current, and the equivalent impedance of the bus after the capacitor is switched is determined in combination with the combination of the capacitor branches to be switched. The equivalent impedance of the bus before switching is compared with the equivalent impedance of the bus after the capacitor is switched on to determine the risk harmonic order, and the harmonic impedance approximation value is determined based on the risk harmonic order.

[0012] Preferably, determining the risk harmonic order and determining the harmonic impedance proximity value based on the risk harmonic order includes: Determine the bus equivalent impedance drop before and after capacitor connection for each harmonic order, and identify the harmonic order whose impedance drop reaches the preset impedance drop judgment limit as the risk harmonic order. Determine the degree of overlap between the risk harmonic order and the harmonic order with the highest proportion, and determine the impedance decrease corresponding to each risk harmonic order; Based on the degree of overlap and the magnitude of impedance drop, the harmonic impedance proximity value is determined, and the risk harmonic order is associated with the corresponding capacitor branch.

[0013] Preferably, in S3, the precursor value of the load sudden change and the close value of the harmonic impedance are comprehensively evaluated, including: Read the conventional compensation demand and its corresponding combination of capacitor branches to be put into operation, and determine the capacity impact level of this switching based on the proportion of the capacity of a single capacitor branch in the conventional compensation demand and the number of capacitor branches to be put into operation. The evaluation proportions of the load mutation precursor value and harmonic impedance proximity value are determined according to the capacity impact level. When the capacity proportion of a single capacitor branch meets the conditions for centralized input, the evaluation proportion of the harmonic impedance proximity value is increased. When multiple branches are input in a decentralized manner, the evaluation proportion of the load mutation precursor value is increased. The load mutation precursor value and harmonic impedance proximity value are synthesized according to the evaluation ratio to obtain the switching risk evaluation result, and the switching risk evaluation result is converted into the switching prediction capability result.

[0014] Preferably, in S3, the prediction capability is divided into a state where the prediction capability is met and a state where the prediction capability is insufficient, based on the prediction capability result, including: Based on whether the low-voltage busbar has records of capacitor branch overcurrent, power factor leading, or harmonic voltage exceeding limits within a preset historical time, a preset judgment limit is determined. The throwing and cutting prediction capability result is compared with the preset judgment limit. When the throwing and cutting prediction capability result reaches the preset judgment limit, it is classified as a prediction capability satisfaction state. When the result of the switching prediction capability does not reach the preset judgment limit, it is classified as a state of insufficient prediction capability, and this state is stored in relation to the current capacitor branch combination to be deployed.

[0015] Preferably, in S4, dynamic adjustment and tiered switching are performed when the predictive capability is insufficient, including: Based on the load mutation precursor value, the single allowable capacity is compressed, and based on the harmonic impedance proximity value and the risk harmonic number, the capacitor branches corresponding to the harmonic risk are eliminated to obtain the adjusted combination of capacitor branches to be put into operation. The adjusted combination of capacitor branches to be put into operation is then switched in stages according to the capacity from small to large, and the waiting time between two adjacent stages is extended based on the load mutation precursor value. After each level of switching is completed, the voltage, current, reactive power and harmonic parameters of the low-voltage bus are collected again, the load change precursor value, harmonic impedance proximity value and switching prediction capability result are updated, and the switching continues, switching is suspended or the most recently connected capacitor branch is disconnected based on the updated switching prediction capability result.

[0016] The technical effects and advantages provided by the present invention in the above technical solution are as follows: The core technical approach of this invention, which distinguishes it from existing technologies, lies in its method of not directly switching capacitor banks after determining the conventional compensation demand. Instead, it further utilizes the short-term variation characteristics of voltage, current, and reactive power to form a precursor value for load abrupt changes, and uses harmonic parameters and the equivalent impedance change after capacitor switching to form a harmonic impedance proximity value. These two factors are then combined to form a switching prediction capability result. Through this process, the controller can identify the trend of impending load start-up / shutdown, impact loading, or periodic fluctuations before switching, and simultaneously identify the risk that the capacitor branch to be switched may cause the impedance of the main harmonic orders to approach each other. This solves the problem of existing technologies relying solely on instantaneous power factor or reactive power and being unable to predict load abrupt changes and harmonic amplification, thus improving the reliability of undercompensation judgment and switching decisions.

[0017] Another core technical approach of this invention lies in dividing the operating state into a state where the predicted capacity is met and a state where the predicted capacity is insufficient, based on the switching prediction capability results. In the state of insufficient capacity, instead of switching all at once according to the conventional compensation demand, the single compensation capacity is compressed based on the precursor value of load sudden changes, and risky capacitor branches are eliminated based on the proximity value of harmonic impedance. Furthermore, a tiered switching process from small to large and post-stage feedback updates are employed. This invention directly avoids overcompensation, voltage rise, harmonic voltage amplification, and capacitor branch overcurrent caused by large-capacity erroneous switching in complex low-voltage power distribution scenarios. It also reduces frequent switching and protection malfunctions, enabling the compensation device to maintain a stable and safe reactive power compensation effect even when load fluctuations and harmonic interference coexist. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0019] Figure 1 This is a flowchart of an adaptive switching control method for a low-voltage complete reactive power compensation device according to the present invention.

[0020] Figure 2 This is a flowchart of the method for generating the prediction results of the cutting and throwing capabilities of the present invention. Detailed Implementation

[0021] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0022] Example 1, please refer to Figure 1 As shown in this embodiment, an adaptive switching control method for a low-voltage complete reactive power compensation device includes: S1: Collect the voltage, current, power factor, reactive power, harmonic parameters, and the capacity of the capacitors already in operation of the low-voltage bus, and determine the conventional compensation requirement based on the power factor.

[0023] After the low-voltage complete reactive power compensation device is put into operation, the controller first synchronously collects the low-voltage bus voltage, current, power factor, reactive power, harmonic parameters, and the capacity of the capacitors already in operation. To avoid the impact of inrush current and voltage transients during capacitor bank switching on the compensation judgment, the sampling process uses the in-phase zero-crossing moment of the low-voltage bus voltage as the synchronous sampling reference. The moment when the low-voltage bus voltage enters the positive half-cycle from the negative half-cycle and the voltage amplitude crosses zero is recorded as a sampling starting point, and then integer cycles are extracted according to the power frequency period. When the low-voltage distribution frequency is 50 Hz, one power frequency period is 20 milliseconds, and the controller can continuously extract voltage and current sampling values ​​within 5 to 10 complete power frequency periods to form the sampling data to be processed.

[0024] During the whole-cycle sampling process, if any capacitor bank is detected to be switched on or off within the sampling interval, data for at least three complete power frequency cycles are discarded from the moment the switching action occurs. If the rate of change of current in the power frequency cycle preceding the switching action exceeds a set allowable value, data for one more complete power frequency cycle is discarded. The remaining sampled data after discarding are reassembled in chronological order, retaining only the data segments with continuous voltage phase and free from switching transients to obtain steady-state sampled data. Through this processing, the subsequent power factor calculation is unaffected by capacitor charging inrush current, contactor jitter, and thyristor conduction transients.

[0025] After obtaining the steady-state sampling data, the fundamental frequency quantity is extracted from the voltage and current sampling values ​​in the steady-state sampling data. Within one complete power frequency cycle, assuming the number of voltage sampling points is N, the i-th voltage sampling value is... The i-th current sample value is Then the effective values ​​of the fundamental voltage and the fundamental current are respectively: ;in, Indicates the effective value of the fundamental voltage. This represents the effective value of the fundamental current. The phase difference between the fundamental voltage and the fundamental current is determined by the difference in their in-phase zero-crossing times. First, the in-phase zero-crossing times of the fundamental voltage are recorded, then the in-phase zero-crossing times of the fundamental current are recorded. The ratio of the time difference between these two times to one power frequency cycle multiplied by 360 degrees gives the fundamental phase difference. The effective power factor used for compensation judgment is the cosine of the fundamental phase difference, and it is only included in subsequent calculations when the effective value of the fundamental voltage is within the allowable range of the rated voltage. Harmonic parameters are determined by the effective values ​​of the voltages and currents of each harmonic other than the fundamental in the steady-state sampling data, and are used to identify whether there is non-fundamental interference in the sampling data.

[0026] The effective power factor is compared with a preset target power factor. The preset target power factor can be a fixed value between 0.95 and 0.99, preferably 0.98. When the effective power factor is lower than the preset target power factor, and the reactive power is inductive, it is determined that the low-voltage bus has a conventional compensation requirement. The current active power is obtained by multiplying the effective value of the fundamental voltage, the effective value of the fundamental current, and the effective power factor; the reactive power allowed to be retained under the target operating condition is obtained by multiplying the current active power by the tangent of the target phase angle, where the target phase angle is the inverse cosine angle corresponding to the preset target power factor; the difference between the current reactive power and the reactive power allowed to be retained under the target operating condition is the inductive reactive power capacity that needs to be reduced. This is then corrected by subtracting the portion of the already invested capacitor capacity that is still effective for the current bus from the inductive reactive power capacity that needs to be reduced, thus obtaining the current capacity deficit that still needs to be compensated.

[0027] After determining the capacity deficit, the controller matches the capacity deficit with the rated capacity of each capacitor branch. During matching, priority is given to capacitor branch combinations that, upon activation, bring the bus power factor close to the preset target power factor without becoming leading. When multiple capacitor branch combinations meet the requirements, priority is given to the combination with the lowest cumulative switching count. When the cumulative switching count is the same, priority is given to the combination with the smallest absolute value of the difference between the single-stage capacity and the capacity deficit. The sum of the rated capacities corresponding to the selected capacitor branch combinations is used as the regular compensation requirement. Therefore, the regular compensation requirement is not directly calculated from the instantaneous power factor, but rather formed after synchronous sampling at zero crossings, integer cycle truncation, transient fluctuation range elimination, determination of the fundamental effective power factor, and correction of the already activated capacitor capacity. This improves the accuracy of undercompensation judgment and reduces erroneous switching caused by transient data in the calculation.

[0028] S2 generates a load surge precursor value based on the short-time variation characteristics of voltage, current and reactive power, and generates a harmonic impedance approximation value based on the harmonic parameters and the equivalent impedance change after the capacitor is connected.

[0029] In this embodiment, the process for determining the precursor value of load sudden change is as follows. The controller uses eight consecutive in-phase zero-crossing cycles as a short-time observation interval. Within each cycle, the effective voltage value, effective current value, and reactive power of the low-voltage bus are arranged in the order of acquisition time to form short-time voltage sequence, short-time current sequence, and short-time reactive power sequence, respectively. The short-time observation interval is preferably eight cycles, corresponding to 160 milliseconds under 50 Hz low-voltage power distribution conditions; when the load start-stop frequency exceeds 20 times per minute, the short-time observation interval is six cycles; when the load start-stop frequency does not exceed 5 times per minute, the short-time observation interval is ten cycles.

[0030] In short-time voltage sequences, the voltage drop between two adjacent voltage cycles is determined by subtracting the effective value of the voltage of the next cycle from the effective value of the previous cycle, dividing by the rated voltage, and then multiplying by 100%. In short-time current sequences, the current rise between two adjacent voltage cycles is determined by subtracting the effective value of the current of the previous cycle from the effective value of the current of the next cycle, dividing by the rated current, and then multiplying by 100%. In short-time reactive power sequences, the continuous increase in inductive reactive power is determined by subtracting the initial inductive reactive power from the inductive reactive power at the end of the continuous increase interval, dividing by the rated compensation capacity of the compensation device, and then multiplying by 100%. A voltage drop of 1.5% is recorded as a voltage disturbance, a current rise of 8% is recorded as a current surge, and a continuous increase in inductive reactive power of 10% is recorded as a reactive power climb. The above three judgment limits are set according to the rated voltage, rated current and rated compensation capacity, with preferred values ​​of 1.5%, 8% and 10% respectively; when the allowable deviation of the power supply voltage is ±7% of the rated voltage, the judgment limit for the voltage drop amplitude is one-fifth to one-quarter of the absolute value of the allowable deviation.

[0031] After confirming the voltage disturbance, current surge, and reactive power rise, they are correlated according to their occurrence frequency. If the frequency of the voltage disturbance occurs earlier than that of the current surge, and the current surge occurs earlier than that of the reactive power rise, and the interval between these two phenomena does not exceed two frequencies, then a load energizing precursor is confirmed. If the frequency correlation is not met, the load energizing precursor value is treated as 0. If the frequency correlation is met, the load energizing precursor value is determined on a percentage basis: voltage disturbance multiplied by 0.30, current surge multiplied by 0.35, and reactive power rise multiplied by 0.35, then summed, and finally overlaid with a duration frequency correction value and a synchronization enhancement correction value. The duration frequency correction value is determined by dividing the duration frequency of the load energizing precursor by the number of frequencies in the short-time observation interval and then multiplying by 20. The synchronization enhancement correction value is determined by dividing the number of times the voltage disturbance, current surge, and reactive power rise simultaneously change in the enhancing direction within adjacent frequencies by the number of frequencies in the short-time observation interval and then multiplying by 20. If the final result exceeds 100, it is counted as 100, and the load mutation precursor value is obtained.

[0032] In this embodiment, the process for determining the equivalent harmonic impedance is as follows: After selecting the combination of capacitor branches to be added based on the conventional compensation requirements, the controller extracts the effective values ​​of each harmonic voltage and each harmonic current from the harmonic parameters. The overall proportion of each harmonic is determined by multiplying the percentage of the effective value of the harmonic voltage relative to the effective value of the fundamental voltage by 0.6, and adding the percentage of the effective value of the harmonic current relative to the effective value of the fundamental current by 0.4. The overall proportions are sorted from high to low, and the top three harmonic orders with an overall proportion of not less than 1% are taken as the harmonic orders with the highest proportions; if there are fewer than three, the actual number is used. Harmonic orders with an effective value of harmonic current less than 0.5% of the rated current are not included in the determination of equivalent impedance.

[0033] The equivalent impedance of the busbar before switching for each harmonic order is determined by dividing the effective value of the harmonic voltage by the effective value of the harmonic current. The harmonic impedance of the branch to be connected at each harmonic order is determined by first dividing the square of the rated voltage by the rated capacity to determine the fundamental capacitive reactance, and then dividing the fundamental capacitive reactance by the corresponding harmonic order. If a reactor is connected in series with the capacitor branch, the fundamental reactance of the reactor is multiplied by the corresponding harmonic order, and then the difference is taken from the capacitive reactance of the capacitor at that harmonic order, and the absolute value is obtained. The equivalent impedance of the busbar after the capacitor is connected is determined by connecting the equivalent impedance of the busbar before switching in parallel with the harmonic impedance of the branch to be connected.

[0034] The equivalent impedance of the busbar before switching is compared with the equivalent impedance of the busbar after the capacitor is connected. The impedance drop is determined by subtracting the equivalent impedance of the busbar after the capacitor is connected from the equivalent impedance of the busbar before switching, dividing by the equivalent impedance of the busbar before switching, and multiplying by 100%. The harmonic number with an impedance drop of 30% is determined as the risk harmonic number; if there is a record of overcurrent in the capacitor branch or harmonic voltage exceeding the limit within the past 30 days, the threshold for judging the impedance drop is 20%; if there is no relevant record, it is 30%.

[0035] Harmonic impedance proximity values ​​are determined on a percentage basis. First, the degree of overlap between the risky harmonic order and the most prevalent harmonic orders is calculated. The overlap is equal to the number of overlapping harmonics divided by the number of the most prevalent harmonics, then multiplied by 100%. Next, the average impedance drop for each risky harmonic order is calculated. The harmonic impedance proximity value is determined by multiplying the overlap by 0.55 and then adding the average impedance drop multiplied by 0.45; if the result exceeds 100, it is rounded down to 100. Through this process, the precursory signs of an impending load energization can be identified before the capacitor bank is put into operation, and the impedance proximity risk to major harmonic orders after capacitor energization can be identified simultaneously, providing a directly usable quantitative basis for subsequent switching prediction capabilities.

[0036] Please see Figure 2As shown in S3, the load mutation precursor value and harmonic impedance proximity value are comprehensively evaluated to obtain the switching prediction capability result, and the prediction capability is divided into a state of sufficient prediction capability and a state of insufficient prediction capability based on the result.

[0037] After obtaining the load surge precursor value, harmonic impedance proximity value, and conventional compensation requirement, the capacity impact level is first determined based on the combination of capacitor branches to be added corresponding to the conventional compensation requirement. If, in the combination of capacitor branches to be added, the capacity of a single capacitor branch accounts for 70% of the conventional compensation requirement, the switching is determined to be at the concentrated capacity impact level; if the capacity of a single capacitor branch accounts for less than 70% of the conventional compensation requirement, and the number of capacitor branches to be added is no less than two, the switching is determined to be at the dispersed capacity impact level; if, in the combination of capacitor branches to be added, there is a situation where the capacity of a single capacitor branch accounts for 50% but less than 70%, the switching is determined to be at the transitional capacity impact level. The above proportions are determined by dividing the rated capacity of a single capacitor branch by the conventional compensation requirement and then multiplying by 100%.

[0038] After determining the capacity impact level, evaluation weights are assigned to the load surge precursor value and harmonic impedance proximity value. Under the concentrated capacity impact level, the evaluation weight for the harmonic impedance proximity value is 0.65, and the evaluation weight for the load surge precursor value is 0.35, to highlight the impact of a single capacitor branch's connection on the impedance drop of a specific harmonic order. Under the dispersed capacity impact level, the evaluation weight for the load surge precursor value is 0.60, and the evaluation weight for the harmonic impedance proximity value is 0.40, to highlight the impact of load changes on compensation capacity matching during the phased connection of multiple branches. Under the transitional capacity impact level, the evaluation weights for both the load surge precursor value and the harmonic impedance proximity value are 0.50. The sum of all evaluation weights is 1.

[0039] After determining the evaluation ratio, the switching risk assessment result is calculated. The calculation formula is: Switching Risk Assessment Result = Load Sudden Change Precursor Value × Load Sudden Change Evaluation Ratio + Harmonic Impedance Approach Value × Harmonic Impedance Evaluation Ratio. Both the load sudden change precursor value and the harmonic impedance approach value are calculated from 0 to 100, therefore the switching risk assessment result is also calculated from 0 to 100. The switching risk assessment result is used to characterize the probability of mis-switching, overcompensation, or harmonic amplification when directly switching according to the conventional compensation demand. Subsequently, the switching risk assessment result is converted inversely into a switching prediction capability result, calculated as: Switching Prediction Capability Result = 100 - Switching Risk Assessment Result. When the switching risk assessment result is 0, the switching prediction capability result is 100; when the switching risk assessment result is 100, the switching prediction capability result is 0.

[0040] After the switching prediction capability result is calculated, it is compared with the preset judgment limit. The preset judgment limit is preferably 60; when there is a record of overcurrent in the capacitor branch, leading power factor, or harmonic voltage exceeding the limit on the low-voltage bus within the past 30 days, the preset judgment limit is 70; when there are no such records within the past 30 days, and the effective voltage value is within the range of 95% to 105% of the rated voltage for 24 consecutive hours, the preset judgment limit is 60. When the switching prediction capability result reaches the preset judgment limit, it is classified as a prediction capability satisfied state, and the step of switching the capacitor bank according to the conventional compensation demand is allowed; when the switching prediction capability result does not reach the preset judgment limit, it is classified as a prediction capability insufficient state, and the switching is not directly performed according to the conventional compensation demand, but enters the subsequent dynamic adjustment step. Therefore, the load change precursor value and harmonic impedance proximity value are no longer used in isolation for judgment. Instead, their evaluation proportion changes with the capacity distribution of the capacitor branch combination to be put into operation. This allows for the priority suppression of harmonic impedance proximity risk when a single branch is put into operation with a large capacity, and the priority suppression of capacity mismatch risk caused by load change when multiple branches are put into operation in a dispersed manner. This improves the correspondence between the switching prediction capability and the actual switching risk.

[0041] S4. When the predicted capacity is met, the capacitor bank is switched on and off according to the conventional compensation requirement. When the predicted capacity is insufficient, the compensation capacity, switching interval and capacitor branch are dynamically adjusted according to the load change precursor value and harmonic impedance proximity value, and then switched in stages. After each stage of switching, the switching predicted capacity result is updated.

[0042] After the switching prediction capability results are divided, capacitor bank switching control is executed. When the switching prediction capability results reach the preset judgment limit and are classified as a state where the prediction capability is met, the controller reads the normal compensation demand and the rated capacity of each capacitor branch, and selects the capacitor branch combination with the smallest absolute value of the difference between the total capacity and the normal compensation demand as the execution switching combination. If there are more than two combinations with the same capacity difference, the combination with the lower cumulative switching count is selected first; if the cumulative switching count is the same, the combination with the most recent exit time being more than 300 seconds earlier is selected first to ensure that the capacitors are discharged completely. After the execution switching combination is determined, they are put into operation in order of increasing capacity; if the execution switching combination contains only one capacitor branch, the operation command is issued directly.

[0043] When the switching prediction capability result fails to reach the preset judgment limit and is classified as insufficient prediction capability, the switching is not directly based on the normal compensation demand. Instead, the single-time allowable switching capacity is determined first. The single-time allowable switching capacity is obtained by subtracting the capacity compression amount from the normal compensation demand. The capacity compression amount is equal to the normal compensation demand multiplied by the load surge precursor value, divided by 100, and then multiplied by the capacity compression coefficient. The capacity compression coefficient is preferably 0.50; when the load surge precursor value reaches 70, the capacity compression coefficient is 0.60; when the load surge precursor value is below 40, the capacity compression coefficient is 0.35. When the single-time allowable switching capacity is lower than the rated capacity of the minimum capacitor branch, the single-time allowable switching capacity is determined according to the rated capacity of the minimum capacitor branch; when the single-time allowable switching capacity is higher than the normal compensation demand, it is determined according to the normal compensation demand.

[0044] After determining the permissible capacity for a single deployment, capacitor branches are eliminated based on harmonic impedance proximity values ​​and risk harmonic orders. Each capacitor branch to be deployed is re-verified based on the change in equivalent bus impedance after deployment. If, after deployment, the impedance drop of a capacitor branch reaches 30% at any risk harmonic order, and that risk harmonic order is among the most prevalent, the capacitor branch is eliminated from the deployment pool. When the harmonic impedance proximity value reaches 70%, the elimination limit for impedance drop is adjusted from 30% to 25%; when the harmonic impedance proximity value is below 50%, the elimination limit remains at 30%. The remaining capacitor branches are accumulated according to their rated capacity until the accumulated capacity does not exceed the permissible capacity for a single deployment and approaches the conventional compensation requirement, resulting in the adjusted combination of capacitor branches to be deployed.

[0045] After the adjusted capacitor branch combination is formed, it is switched on in stages according to capacity from smallest to largest. A waiting time is set between two adjacent stages of switching, which is obtained by adding the basic waiting time and the precursor delay time. The basic waiting time is preferably 60 seconds; the precursor delay time is equal to the load change precursor value divided by 100 and then multiplied by 120 seconds. Thus, when the load change precursor value is 50, the precursor delay time is 60 seconds and the waiting time is 120 seconds; when the load change precursor value is 80, the precursor delay time is 96 seconds and the waiting time is 156 seconds. The upper limit of the waiting time is set to 240 seconds and the lower limit is set to 60 seconds. If the power factor has reached the preset target power factor during the waiting period, or the reactive power changes from inductive to capacitive, the subsequent stage of switching is canceled.

[0046] After each level of switching is completed, the voltage, current, reactive power, and harmonic parameters of the low-voltage bus are re-acquired, and the load surge precursor value, harmonic impedance proximity value, and switching prediction capability results are updated in the aforementioned manner. If the updated switching prediction capability results reach the preset judgment limit and the reactive power still exhibits inductive reactive power, the next level of switching is executed. If the updated switching prediction capability results are lower than the preset judgment limit but not lower than the preset judgment limit by 10 points, subsequent levels of switching are suspended, and the short-time observation interval is re-entered for data acquisition. If the updated switching prediction capability results are lower than the preset judgment limit by more than 10 points, or if any of the following conditions are met after switching: the power factor becomes leading, the harmonic impedance proximity value reaches 80, or the capacitor branch current reaches 1.25 times the rated current, the most recently switched capacitor branch is disconnected. Through this processing method, the compensation capacity, switching interval, and capacitor branch selection under insufficient prediction capability conditions can be gradually corrected according to the real-time operating status, avoiding overcompensation, harmonic amplification, and capacitor branch overcurrent caused by one-time switching.

[0047] Example 2 verifies the switching stability and harmonic suppression effect of an adaptive switching control method for a low-voltage complete reactive power compensation device in a complex low-voltage power distribution scenario. The test object is a 0.4 kV low-voltage power distribution circuit with a 1000 kVA distribution transformer and a rated current of 1443 A. The rated compensation capacity of the reactive power compensation device is 360 kVAR, and the rated capacities of the capacitor branches are 20 kVAR, 20 kVAR, 30 kVAR, 30 kVAR, 40 kVAR, 40 kVAR, 50 kVAR, 60 kVAR, and 70 kVAR, respectively. Each capacitor branch has independent switching control. The loads connected to the low-voltage bus include two 75 kW variable frequency fans, one 110 kW air compressor, three spot welding machines, one rectifier power supply, and general lighting and small power loads. The variable frequency fans and spot welding machines will cause 5th and 7th harmonic current fluctuations, and the start-up and shutdown of the air compressor will cause a short-term surge in inductive reactive power.

[0048] The experimental setup compared the comparative and implementation examples. The comparative example used a switching method based solely on the instantaneous power factor to determine the compensation capacity. When the power factor was below 0.95, the capacitor branch was directly connected according to the conventional compensation requirement; when the power factor was ahead, the corresponding capacitor branch was directly disconnected. The implementation example used the method described in this application. After obtaining the conventional compensation requirement, it further determined the load surge precursor value, harmonic impedance proximity value, and switching prediction capability results. When the prediction capability was insufficient, the compensation capacity, switching interval, and capacitor branch were dynamically adjusted and switched in stages. The total test duration was 60 minutes, with the first 20 minutes for variable frequency fan speed regulation, the middle 20 minutes for air compressor start-up and shutdown and concentrated operation of spot welding equipment, and the last 20 minutes for load reduction. Each control method was repeated three times, and the data in the table is the average of the three tests.

[0049] Table 1. Operational data for three typical time periods during the experiment and key evaluation results generated by the method of this application.

[0050] As shown in Table 1, during the variable frequency fan speed regulation stage, although the power factor is 0.88 and a 160 kV capacitor is required according to the conventional compensation demand, the harmonic impedance approaches 48, and the switching prediction capability result is 58, which is lower than the preset judgment limit of 60. Therefore, the method in this application does not directly input 160 kV, but instead enters dynamic adjustment. The controller first compresses the single allowable input capacity and eliminates the 70 kV capacitor branch that causes an impedance drop of up to 30% at the 5th harmonic. Finally, 20 kV, 30 kV, and 40 kV branches are selected for tiered input. After each level of switching, voltage, current, reactive power, and harmonic parameters are re-collected. The next level of switching is executed only after confirming that the switching prediction capability result has not continued to decline.

[0051] During the air compressor startup phase, the load surge precursor value reached 76, indicating that voltage disturbances, current surges, and inductive reactive power increases occurred sequentially. If a 220 kvar capacitor were directly connected at this point, overcompensation could easily occur after the air compressor startup was complete. This application's method, based on the load surge precursor value, reduces the allowable single-stage capacity to approximately 62% of the conventional compensation requirement and extends the waiting time between adjacent switching stages from 60 seconds to 151 seconds. In the experiment, after the first 40 kvar capacitor branch was connected, the air compressor entered stable operation. The recalculated load surge precursor value decreased to 42, and the switching prediction capability increased to 61. Subsequently, only a 30 kvar capacitor branch was added, and the remaining large-capacity branches were not connected, thus avoiding capacitive overcompensation after the load stabilized.

[0052] Table 2. Comparison of switching results between the comparative example and the method of this application under the same working conditions.

[0053] Table 2 shows that the comparative example had 47 switching operations within 60 minutes, including 9 erroneous switching operations. These erroneous switching operations manifested as the capacitor branch being switched on after the load decreased, a short-term leading power factor, and an increase in harmonic voltage after the capacitor branch was switched on. The method in this application reduces the total number of switching operations to 23 and the number of erroneous switching operations to 1. This indicates that by using the load abrupt change precursor value and harmonic impedance proximity value in the switching prediction, repeated switching caused by instantaneous power factor fluctuations can be reduced. The comparative example showed a maximum 5th harmonic voltage distortion rate of 5.8% and a maximum capacitor branch current of 1.38 times the rated current. In the method of this application, the maximum 5th harmonic voltage distortion rate was reduced to 3.4%, and the maximum capacitor branch current was reduced to 1.16 times the rated current. This demonstrates that by eliminating capacitor branches corresponding to risky harmonic orders, the possibility of the bus equivalent impedance approaching the main harmonic order after capacitor switching can be reduced.

[0054] Table 3. Data table of the tiered switching process of the method in this application under conditions of insufficient predictive capability.

[0055] As shown in Table 3, when the normal compensation demand is 220 kV, the comparative method would directly put a large-capacity branch combination into operation according to capacity matching. However, the method in this application first compresses the single-time allowable capacity based on the load change precursor value, and then eliminates risky branches based on the harmonic impedance proximity value, ultimately only putting in 70 kV. After the first stage of 40 kV input, the power factor improved from 0.82 to 0.89, but the switching prediction capability result was still 52, which did not reach the preset judgment limit. Therefore, it entered the observation stage instead of immediately putting in a large capacity. After the second stage of 30 kV input, the power factor improved to 0.94, and the switching prediction capability result improved to 61, reaching the preset judgment limit. Subsequently, during the waiting period, it was detected that the power factor continued to recover to 0.96, and the inductive reactive power decreased significantly. The subsequent 80 kV capacitor branch was canceled from being put into operation. This process demonstrates that the method in this application is not simply a delayed switching, but rather a re-determination of subsequent actions after each switching stage using updated load mutation precursor values, harmonic impedance proximity values, and switching prediction capabilities.

[0056] The experimental data above demonstrates that the method described in this application achieves the following beneficial effects in complex low-voltage power distribution scenarios: First, by determining the precursor value of load abrupt changes through short-time variation characteristics, the single-time input capacity can be reduced when impact loads such as air compressors and spot welding equipment are about to change their operating state, thus avoiding overcompensation after the load stabilizes. Second, by determining the harmonic impedance proximity value through harmonic parameters and the equivalent impedance change after capacitor input, risky capacitor branches corresponding to the main harmonic order can be identified before capacitor bank input, reducing harmonic voltage amplification and capacitor branch overcurrent. Third, by classifying the prediction capability into a state of sufficient prediction capability and a state of insufficient prediction capability based on the switching prediction capability results, the conventional compensation demand can be directly executed only when operating conditions permit, and switched to dynamic adjustment when operating conditions are unfavorable. Fourth, through tiered switching and post-stage feedback updates, the number of switching operations can be reduced, the power factor lead time can be shortened, and the average power factor after compensation can be improved. The above data demonstrates that, compared to methods that rely solely on instantaneous power factor for switching, the method of this application can significantly improve the switching accuracy, operational stability, and capacitor branch safety of low-voltage reactive power compensation devices under complex load conditions.

[0057] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

Claims

1. An adaptive switching control method for a low-voltage complete reactive power compensation device, characterized in that, include: S1. Collect the voltage, current, power factor, reactive power, harmonic parameters and the capacity of the capacitors already in operation of the low-voltage bus, and determine the conventional compensation requirement based on the power factor. S2, generates load surge precursor values ​​based on the short-time variation characteristics of voltage, current and reactive power, and generates harmonic impedance approximation values ​​based on the harmonic parameters and the equivalent impedance change after capacitor connection. S3, comprehensively evaluate the load change precursor value and harmonic impedance proximity value to obtain the switching prediction capability result, and classify the prediction capability as satisfied or insufficient based on the result; S4. When the predicted capacity is met, the capacitor bank is switched on and off according to the conventional compensation requirement. When the predicted capacity is insufficient, the compensation capacity, switching interval and capacitor branch are dynamically adjusted according to the load change precursor value and harmonic impedance proximity value, and then switched in stages. After each stage of switching, the switching predicted capacity result is updated.

2. The adaptive switching control method for a low-voltage complete reactive power compensation device according to claim 1, characterized in that, Determining the conventional compensation requirement based on the power factor includes: Using the in-phase zero-crossing moment of the low-voltage bus voltage as the sampling reference, the voltage and current are subjected to integer truncation, and the transient fluctuation range after the capacitor bank is put into or removed is eliminated to obtain steady-state sampling data. The fundamental voltage, fundamental current, and effective power factor used for compensation judgment are determined based on the steady-state sampling data. The effective power factor is compared with the preset target power factor, and the capacity deficit is determined by combining the reactive power and the capacity of the capacitors already in operation. The capacity deficit is then matched with the capacity of the capacitor branches to obtain the conventional compensation requirement.

3. The adaptive switching control method for a low-voltage complete reactive power compensation device according to claim 2, characterized in that, The transient fluctuation range after the capacitor bank is connected or disconnected is excluded, including: Record the connection or disconnection time of each capacitor branch, and simultaneously record the start and end cycles of the low-voltage bus current jump. Based on the cycle in which the input or cut-off time occurs, cycles containing current jumps and their adjacent transition cycles are eliminated. The phase-continuous integer-cycle data after removal are spliced ​​together in chronological order to serve as steady-state sampling data for determining the effective power factor and harmonic parameters.

4. The adaptive switching control method for a low-voltage complete reactive power compensation device according to claim 3, characterized in that, S2 generates load surge precursor values ​​based on short-time variation characteristics of voltage, current, and reactive power, including: Using multiple consecutive in-phase zero-crossing cycles as short-time observation intervals, the effective voltage, effective current, and reactive power of the low-voltage bus are sorted cycle by cycle to obtain short-time voltage, short-time current, and short-time reactive power sequences. Based on the short-time voltage sequence, short-time current sequence, and short-time reactive power sequence, the voltage disturbance, current surge, and reactive power rise are determined respectively. The voltage disturbance, current surge, and reactive power rise are correlated in chronological order of occurrence, and the load surge precursor value is determined based on the correlation results and the number of continuous cycles.

5. The adaptive switching control method for a low-voltage complete reactive power compensation device according to claim 4, characterized in that, The load mutation precursor value is determined based on the correlation results and the number of sustained cycles, including: When the frequency corresponding to the voltage disturbance is earlier than the frequency corresponding to the current surge, and the frequency corresponding to the current surge is earlier than the frequency corresponding to the reactive power rise, it is confirmed that there is a precursor to load energization. The number of continuous cycles of the load energizing precursor within a short observation interval is counted, and the number of times the voltage disturbance, current surge, and reactive power rise synchronously increase in adjacent cycles is determined. Based on the number of continuous cycles and the number of synchronous enhancements, the voltage disturbance, current surge, and reactive power rise are corrected to obtain the load change precursor value.

6. The adaptive switching control method for a low-voltage complete reactive power compensation device according to claim 5, characterized in that, S2 generates harmonic impedance approximation values ​​based on the harmonic parameters and the equivalent impedance change after capacitor connection, including: Based on the conventional compensation requirements, the combination of capacitor branches to be put into operation is determined, and the harmonic order, harmonic voltage and harmonic current with the highest proportion at the low voltage bus are extracted from the harmonic parameters. The equivalent impedance of the bus before switching is determined based on the harmonic voltage and harmonic current, and the equivalent impedance of the bus after the capacitor is switched is determined in combination with the combination of the capacitor branches to be switched. The equivalent impedance of the bus before switching is compared with the equivalent impedance of the bus after the capacitor is switched on to determine the risk harmonic order, and the harmonic impedance approximation value is determined based on the risk harmonic order.

7. The adaptive switching control method for a low-voltage complete reactive power compensation device according to claim 6, characterized in that, Determining the risk harmonic order and, based on the risk harmonic order, determining the approximate harmonic impedance value, including: Determine the bus equivalent impedance drop before and after capacitor connection for each harmonic order, and identify the harmonic order whose impedance drop reaches the preset impedance drop judgment limit as the risk harmonic order. Determine the degree of overlap between the risk harmonic order and the harmonic order with the highest proportion, and determine the impedance decrease corresponding to each risk harmonic order; Based on the degree of overlap and the magnitude of impedance drop, the harmonic impedance proximity value is determined, and the risk harmonic order is associated with the corresponding capacitor branch.

8. The adaptive switching control method for a low-voltage complete reactive power compensation device according to claim 7, characterized in that, S3 comprehensively evaluates the load mutation precursor value and harmonic impedance proximity value, including: Read the conventional compensation demand and its corresponding combination of capacitor branches to be put into operation, and determine the capacity impact level of this switching based on the proportion of the capacity of a single capacitor branch in the conventional compensation demand and the number of capacitor branches to be put into operation. The evaluation proportions of the load mutation precursor value and harmonic impedance proximity value are determined according to the capacity impact level. When the capacity proportion of a single capacitor branch meets the conditions for centralized input, the evaluation proportion of the harmonic impedance proximity value is increased. When multiple branches are input in a decentralized manner, the evaluation proportion of the load mutation precursor value is increased. The load mutation precursor value and harmonic impedance proximity value are synthesized according to the evaluation ratio to obtain the switching risk evaluation result, and the switching risk evaluation result is converted into the switching prediction capability result.

9. The adaptive switching control method for a low-voltage complete reactive power compensation device according to claim 8, characterized in that, S3 classifies the prediction capability into two states based on the prediction capability result: a prediction capability satisfied state and a prediction capability insufficient state, including: Based on whether the low-voltage busbar has records of capacitor branch overcurrent, power factor leading, or harmonic voltage exceeding limits within a preset historical time, a preset judgment limit is determined. The throwing and cutting prediction capability result is compared with the preset judgment limit. When the throwing and cutting prediction capability result reaches the preset judgment limit, it is classified as a prediction capability satisfaction state. When the switching prediction result does not reach the preset judgment limit, it is classified as a state of insufficient prediction capability, and this state is stored in relation to the current capacitor branch combination to be deployed.

10. The adaptive switching control method for a low-voltage complete reactive power compensation device according to claim 9, characterized in that, In S4, dynamic adjustments are made to the tiered deployment and switching when the predictive capability is insufficient, including: Based on the load mutation precursor value, the single allowable capacity is compressed, and based on the harmonic impedance proximity value and the risk harmonic number, the capacitor branches corresponding to the harmonic risk are eliminated to obtain the adjusted combination of capacitor branches to be put into operation. The adjusted combination of capacitor branches to be put into operation is then switched in stages according to the capacity from small to large, and the waiting time between two adjacent stages is extended based on the load mutation precursor value. After each level of switching is completed, the voltage, current, reactive power and harmonic parameters of the low-voltage bus are collected again, the load change precursor value, harmonic impedance proximity value and switching prediction capability result are updated, and the switching continues, switching is suspended or the most recently connected capacitor branch is disconnected based on the updated switching prediction capability result.

Images