A regional power grid power quality treatment method for multi-type subject access

Abstract

The application belongs to the cross field of electric power system and power electronics technology, and particularly relates to a regional power grid power quality treatment method for multi-type subject access. By introducing advanced phase compensation, the phase lag of the detection link is calculated and fully compensated in real time, so that the extracted harmonic component is instantaneously synchronized with the actual harmonic of the power grid. In combination with an adaptive hybrid modulation technology, the dominant harmonic frequency band can be focused on compensation. A fully decoupled control based on feedforward is adopted, which eliminates the cross coupling between the d and q axis currents, so that the inner loop controller can realize fast tracking with almost no overshoot. Meanwhile, the adaptive adjustment mechanism of the PLL bandwidth ensures that the synchronization signal is fast and robust when the power grid changes.

Description

Technical Field

[0001] This invention belongs to the interdisciplinary field of power system and power electronics technology, and specifically relates to a comprehensive power quality management method for regional power grids that is suitable for the access of new entities with a high proportion of power electronics (such as distributed photovoltaic, energy storage power stations, electric vehicle charging stations, data centers, etc.). Background Technology

[0002] With the increasing number of new types of entities connecting to the power grid, the current regional power grid power quality management faces three core challenges: Diversification and widebanding of harmonic sources: Traditional single-type harmonic sources (such as industrial rectifier loads) have been replaced by new entities with diverse characteristics and wide harmonic frequencies, making them difficult to effectively handle with traditional passive filters and fixed-tuned APFs. Intensified power fluctuations: The intermittency of photovoltaic and wind power generation and the randomness of electric vehicle charging lead to rapid changes in the direction and magnitude of power flow in the grid, requiring management equipment to have bidirectional, fast, and accurate active / reactive current detection and compensation capabilities. Isolated management devices: Existing management devices (such as SVG and APF) are mostly designed and operated independently, lacking coordination mechanisms. When multiple devices are connected in parallel, problems such as circulating current, oscillation, and compensation conflicts easily arise, resulting in poor system scalability and preventing optimal overall efficiency and reliability.

[0003] Existing power management methods mostly focus on topology optimization or control algorithm improvement of single devices, lacking a systematic solution that takes a holistic view of the regional power grid and coordinates the control and resource scheduling of multiple management devices as an organic whole. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides a regional power grid power quality management method for multiple types of entities accessing the grid. This method integrates various management devices as an organic whole for coordinated control and resource scheduling, which can significantly improve management accuracy and ensure dynamic response speed.

[0005] To solve the aforementioned technical problem, the technical solution adopted by the present invention is: a method for improving power quality in a regional power grid for multiple types of entities accessing the network, comprising the following steps: S01. Sample the three-phase grid voltage and three-phase grid current. Transform the sampled data to a two-phase rotating dq coordinate system using Clark and Park transformations. After Clark transformation, the current in the αβ coordinate system is i_α and i_β. After Park transformation, the direct-axis component of the voltage in the dq coordinate system is U_d and the quadrature-axis component is U_q. The direct-axis component of the current is i_d and the quadrature-axis component is i_q. After performing phase lead compensation on i_α and i_β, Park transformation is performed to obtain the fundamental current components i_d_fund and i_q_fund without delay. S02. Calculate the effective voltage deviation ΔU and the total harmonic distortion (THD) of the current. The global mode command, along with ΔU and THD, is input to the local decision-maker. The local decision-maker dynamically generates the harmonic compensation weighting coefficient K_h and the fundamental reactive current command i_q_svg_ref based on the preset priority logic. When the global mode instruction is voltage regulation priority mode and |ΔU| is greater than the set threshold, the local decision-maker increases i_q_svg_ref and decreases K_h, so that the governance module can give priority to the SVG function to support the voltage. When the global mode instruction is harmonic suppression priority mode and THD is greater than the set threshold, the local decision-maker increases K_h and decreases i_q_svg_ref, so that the governance module can give priority to the APF function to suppress harmonics. S03. The governance module adopts dual closed-loop control, with an outer loop being a voltage loop and an inner loop being a current loop. The voltage outer loop contains two independently operating outer loop controllers: a DC voltage outer loop and an AC voltage outer loop. The DC voltage outer loop outputs the fundamental active current d-axis component i_d_dc_ref; the AC voltage outer loop outputs the fundamental reactive current q-axis component i_q_fund_ref. i_d_dc_ref and i_q_fund_ref are synthesized with harmonic commands and then corrected by thermal equalization to obtain the final inner loop commands i_d_ref and i_qref. S04. Generate voltage commands u_d_ref and u_q_ref based on the inner loop command: , in ω is the gain coefficient of the current loop, L_f is the fundamental angular frequency of the power grid, and L_f is the output filter inductance value of the inverter. S05. Perform Park and Clark inverse transformations on the voltage command to obtain a three-phase modulation wave. Combine this with a dynamic modulation strategy to generate a PWM signal. Drive the converter to output compensation current based on the PWM signal.

[0006] Furthermore, the thermal balance correction in step S03 is as follows: all governance modules communicate with each other, sending data frames including heatsink temperatures, calculating the average temperature of all governance modules, and obtaining the final inner loop instructions i_d_ref and i_q_ref based on the average temperature. , This is the final inner loop instruction for the i-th current loop. The inner loop instruction before correction is defined as follows: T_j_i is the heatsink temperature of the i-th governance module, T_avg is the average heatsink temperature of all governance modules, and K is the temperature-impedance adjustment coefficient.

[0007] Furthermore, the local decision-maker dynamically generates harmonic compensation weighting coefficient K_h and fundamental reactive current command i_q_svg_ref based on the global mode command, ΔU, and THD. K_h is used to generate the harmonic compensation command, and i_q_svg_ref is used to adjust the voltage reference value of the AC voltage outer loop. , in , This is the command for d-axis and q-axis harmonic compensation current. , These are d-axis and q-axis harmonic currents with no time delay. , U_ref=Unom+Ku iq_svg_ref, Wherein, U_ref is the voltage reference value of the outer loop of AC voltage, that is, the voltage target that the PCC point is expected to maintain; Unom is the rated voltage, that is, the nominal voltage of the power grid during normal operation; iq_svg_ref is the fundamental reactive current target, reflecting the reactive demand intensity under the current mode; and Ku is the voltage-reactive conversion coefficient, which is the proportional factor that converts the reactive current target into the voltage offset.

[0008] Furthermore, in step S01, the phase delay τ_total for the lead phase compensation is: , The formula for leading phase compensation is: , Where N is the moving average filter window length, k is the sampling time, f_s is the sampling frequency, T_s is the control period; i_α_compensated(k) and i_β_compensated(k) are the α-axis currents after phase compensation, ω is the grid current angular frequency, and i_α(k) and i_β(k) are the α-axis currents before phase compensation.

[0009] Furthermore, in step S02, the voltage RMS deviation ΔU is calculated as follows: the fundamental voltage RMS value is calculated based on the voltage direct-axis component u_d in the dq coordinate system. And thus obtain U nom This is the rated voltage.

[0010] Furthermore, the total harmonic distortion (THD) of the current is calculated as follows: based on the fundamental current components i_d_fund and i_q_fund without delay, the harmonic current components i_d_harm and i_q_harm are calculated. The fundamental current component and harmonic current component without delay are converted into the total effective value of the fundamental current I_fund and the total effective value of the harmonic current I_harm, and then calculated. .

[0011] Furthermore, the AC voltage outer loop control process is as follows: The three-phase voltage at the PCC point is sampled in real time, and after phase-locked loop (PLL) and dq transformation, the d-axis voltage component u_d is obtained. The deviation ΔU between u_d and U_ref is calculated, and the deviation ΔU is fed into the outer loop. The outer loop output i_q_fund_ref is: , Where K_droop is the droop coefficient, Q is the reactive power output of the machine, Q_avg is the average reactive power of the system, and Kp and Ki are the AC proportional gain and AC integral gain of the outer loop. The outer loop output and harmonic compensation command are combined to form the initial base value i_q_ref_original for the inner loop: i_q_ref_original=i_q_fund_ref+i_q_h_ref.

[0012] Furthermore, the DC voltage outer loop control process is as follows: The DC bus voltage U_dc is sampled in real time, and the sampled value is compared with the given voltage reference value U_dc_ref to obtain the error e_dc(t) = U_dc_ref - U_dc(t). This error is then fed into the outer loop, and the outer loop output i_d_dc_ref(t) is: , Where Kp_d and Ki_dc are the DC proportional gain and DC integral gain of the outer loop, respectively. The outer loop output and harmonic compensation command are combined to form the initial base value i_d_ref_original for the inner loop i_q_ref: i_d_ref_original =i_d_dc_ref+i_d_h_ref.

[0013] Furthermore, combining a dynamic modulation strategy, the PWM signal is generated as follows: after obtaining the delay-free fundamental current components i_d_fund and i_q_fund, according to... Harmonic current components are acquired, and FFT analysis is performed on them to identify the dominant harmonic order and its amplitude, thus obtaining the harmonic spectrum. SVPWM modulation is applied to the fundamental and lower harmonic frequency bands. A resonant controller of a specific order is dynamically injected near the identified dominant harmonic frequency band, and the modulation method for this frequency band is switched to DPWM with the optimal switching frequency or specific harmonic elimination PWM. The carrier frequency is dynamically adjusted within the specified harmonic spectrum range based on the heat sink temperature. A PWM signal is generated by combining the voltage signal from step S04 and the modulation method. The PWM signal enters the converter, controlling the power module of the improved ANPC topology to generate the required compensation current, which is finally injected into the power grid to complete power quality management.

[0014] Furthermore, within the specified carrier frequency range, when the radiator temperature is higher than the set value, the carrier frequency is reduced; when the radiator temperature is lower than the set value, the carrier frequency is increased.

[0015] The beneficial effects of this invention are as follows: By introducing advanced phase compensation, the phase lag in the detection process is calculated and fully compensated in real time, enabling the extracted harmonic components to achieve instantaneous synchronization with the actual harmonics of the power grid. This elevates the accuracy of harmonic detection from "approximate with delay" to "precise with no delay." Combined with adaptive hybrid modulation technology, it is possible to perform targeted compensation for the dominant harmonic frequency band.

[0016] Employing a fully decoupled control based on feedforward eliminates the cross-coupling between the d-axis and q-axis currents, enabling the inner-loop controller to achieve near-overshoot-free fast tracking. Simultaneously, the PLL bandwidth adaptive adjustment mechanism ensures both speed and robustness of the synchronization signal despite changes in grid strength.

[0017] The adaptive resonant modulation technique dynamically adjusts the modulation strategy and switching frequency based on the real-time harmonic spectrum. When the harmonic content is low, the switching frequency can be automatically reduced to decrease switching losses; when a specific harmonic exceeds the limit, the optimal modulation mode is used for precise compensation. Attached Figure Description

[0018] Figure 1 This is a flowchart of the method; Figure 2 A waveform comparison of the current before and after compensation with the actual current. Detailed Implementation

[0019] The present invention will be further described below with reference to the accompanying drawings and specific embodiments.

[0020] Example 1 This embodiment discloses a method for improving power quality in regional power grids with multiple types of entities connected, such as... Figure 1 As shown, it includes the following steps: Step 1: Real-time current detection and decomposition based on dynamic delay compensation. This step is fundamental to achieving accurate compensation. Its purpose is to quickly and without delay separate the fundamental active current, fundamental reactive current, and harmonic currents from the complex grid current. The specific steps are as follows: Signal Acquisition and Preprocessing: A high-speed ADC (Analog-to-Digital Converter) is used to synchronously sample the three-phase grid voltages Ua, Ub, and Uc, and the three-phase grid currents i_a, i_b, and i_c, at a sampling frequency of no less than 20kHz. The sampled values ​​are processed for anti-aliasing using a digital filter (first-order low-pass filter, cutoff frequency 2kHz). Voltage and current signals are then normalized.

[0021] Coordinate Transformation and Preliminary Filtering: Using the Clark and Park transformation formulas, the voltage and current in the three-phase stationary coordinate system (abc) are transformed to the dq coordinate system, which rotates synchronously with the fundamental positive-sequence voltage of the power grid. After the transformation, the direct-axis components U_d and U_q of the voltage, and the direct-axis components i_d and i_q of the current are obtained. i_d and i_q are then passed through a moving average filter (MAF). The window length N of the MAF is set to one fundamental period to filter out all integer harmonics.

[0022] The current in the three-phase stationary coordinate system (abc) transformed to the αβ coordinate system using Clark transformation is: , The current in the αβ coordinate system, transformed to the dq coordinate system using the Park transformation, is: .

[0023] Dynamic delay observation and compensation: It is recognized that MAF filtering and digital control calculation introduce a fixed phase delay τ_total. τ_total consists of the group delay of MAF (N / 2 sampling periods, where N is the window length) and the calculation delay (approximately 1.5 control periods).

[0024] a. Delay calculation: The dynamic delay observer calculates τ_total=(N / (2) in real time. f_s))+1.5 T_s, where f_s is the sampling frequency and T_s is the control period.

[0025] b. Lead compensation: Perform lead phase compensation on the currents i_α and i_β in the α and β coordinate systems after Clark transformation. Construct the compensation operator: , i_α_compensated(k) and i_β_compensated(k) are the α-axis currents and β-axis currents after phase compensation, respectively. ω is the angular frequency of the grid current, which is provided in real time by the PLL. i_α(k) and i_β(k) are the α-axis currents and β-axis currents before phase compensation, respectively.

[0026] c. Transform again: Perform Park transformation on the compensated i_α_compensated and i_β_compensated to obtain the fundamental current components i_d_fund and i_q_fund without delay, and calculate the harmonic current: i_d_harm=i_d-i_d_fund, i_q_harm=i_q-i_q_fund.

[0027] Phase compensation completely eliminates the delay in the detection process, ensuring that the extracted fundamental and harmonic components are synchronized with the actual instantaneous value of the grid current, thus providing accurate real-time performance for subsequent control. Figure 2 The waveforms of the current before and after compensation are compared with the actual current. It can be seen that the current waveform after compensation is closer to the actual waveform.

[0028] Step 2: Multi-objective collaborative decision-making and adaptive output impedance generation.

[0029] This step involves autonomously generating governance instructions based on the high-precision power grid status information extracted in step one, and achieving autonomous collaboration and thermal management among governance modules.

[0030] Based on the output of step one, the local controller calculates two key decision variables: Voltage RMS deviation ΔU: The RMS value of the fundamental voltage is calculated using the direct-axis component u_d of the fundamental positive-sequence voltage obtained in step one in a synchronously rotating coordinate system. And thus obtain U nom This is the reference value for the rated voltage, and this variable characterizes the stability of the power grid voltage.

[0031] Total Harmonic Distortion (THD): Using the undelayed fundamental current components i_d_fund and i_q_fund, and the harmonic current components i_d_harm and i_q_harm separated after dynamic delay compensation in step one, the effective value of the fundamental current I_fund and the total effective value of the harmonic current I_harm are calculated respectively. , This leads to the equation THD = (I_harm / I_fund) × 100%. This variable represents the purity of the grid current.

[0032] The system-level monitoring platform issues a global mode command (Mode Word) and inputs it to the local decision unit along with the locally calculated ΔU and THD in real time. The decision unit dynamically generates the harmonic compensation weighting coefficient K_h and the fundamental reactive current command i_q_svg_ref based on the global mode command, ΔU, and THD. K_h is used to generate the harmonic compensation command, and i_q_svg_ref is used to adjust the voltage reference value of the AC voltage outer loop.

[0033] When the global mode instruction is voltage regulation priority mode and |ΔU| is greater than the set threshold, the local decision-maker increases i_q_svg_ref and appropriately decreases K_h, so that the governance module can give priority to the SVG function to support the voltage. When the global mode instruction is in harmonic suppression priority mode and THD is greater than the set threshold, the local decision-maker increases K_h and appropriately decreases i_q_svg_ref, so that the governance module can give priority to the APF function to suppress harmonics.

[0034] The formula for generating harmonic compensation instructions based on K_h is: , in , This is the command for d-axis and q-axis harmonic compensation current. , These are d-axis and q-axis harmonic currents with no time delay. .

[0035] The formula for adjusting the voltage reference value U_ref of the outer loop of AC voltage based on i_q_svg_ref is: U_ref=Unom+Ku iq_svg_ref, Wherein, U_ref is the voltage reference value of the AC voltage outer loop, that is, the voltage target that the PCC point is expected to maintain; Unom is the rated voltage, that is, the nominal voltage of the power grid during normal operation; iq_svg_ref is the fundamental reactive current command, reflecting the reactive demand intensity under the current mode; and Ku is the voltage-reactive conversion coefficient, which is the proportional factor that converts the reactive current target into the voltage offset.

[0036] The governance module is configured with dual closed-loop control, with an outer loop being a voltage loop and an inner loop being a current loop. The outer voltage loop contains two independently operating outer loop controllers: a DC voltage outer loop and an AC voltage outer loop. The DC voltage outer loop outputs the fundamental active current d-axis component i_d_dc_ref; the AC voltage outer loop outputs the fundamental reactive current q-axis component i_q_fund_ref. i_d_dc_ref and i_q_fund_ref are synthesized with harmonic commands and then corrected by thermal equalization to obtain the final inner loop commands i_d_ref and i_q_ref.

[0037] Specifically, the AC voltage outer loop control process is as follows: Sample the three-phase voltage at the PCC point, and after phase-locked loop (PLL) and dq transformation, obtain the d-axis voltage component u_d. Calculate the deviation ΔU between u_d and the rated voltage U_ref. The deviation ΔU is sent to the outer loop, and the outer loop output i_q_fund_ref is: , Where K_droop is the droop coefficient, Q is the reactive power output of the machine, Q_avg is the average reactive power of the system, and Kp and Ki are the AC proportional gain and AC integral gain of the outer loop. The outer loop output and the harmonic compensation command are combined to form the initial base value i_q_ref_original for the inner loop: i_q_ref_original=i_q_fund_ref+i_q_h_ref.

[0038] The DC voltage outer loop control process is as follows: The DC bus voltage U_dc is sampled in real time. The sampled value is compared with the given voltage reference value U_dc_ref to obtain the error e_dc(t) = U_dc_ref - U_dc(t). This error is then fed into the outer loop, and the outer loop output i_d_dc_ref(t) is: , Where Kp_d and Ki_dc are the DC proportional gain and DC integral gain of the outer loop, respectively. The outer loop output and the harmonic compensation command are combined to form the initial base value i_d_ref_original for the inner loop i_q_ref: i_d_ref_original = i_d_dc_ref + i_d_h_ref. The above steps yield the initial inner-loop control commands for the governance module. Considering the individual differences among the multiple governance modules—their operating environments, connected loads, and heatsink temperatures—a thermal equalization correction is performed on the inner-loop control commands based on these individual differences (especially heatsink temperatures). The thermal equalization correction process is as follows: All online governance modules communicate via a low-bandwidth fieldbus network (such as a CAN bus). Each module broadcasts a standardized data frame to the bus at a fixed period (e.g., T_comm = 100ms). This data frame contains at least: module ID, heatsink temperature T_j_i, RMS output current I_out_i, and the module's health status.

[0039] The IGBT heatsink temperature T_j_i is collected in real time by a high-precision temperature sensor embedded in the IGBT heatsink. This temperature directly reflects the thermal stress level and cumulative aging effects of the power module, and is the primary indicator for assessing the module's health status.

[0040] The effective value of the output current I_out_i is measured and calculated by the current sensor on the output side of the module, and is used to evaluate the instantaneous output of the module.

[0041] After receiving status information from other modules, each module performs a consistency calculation locally to form a system-level collaborative baseline: Calculation of average temperature T_avg: T_avg=(ΣT_i) / M, Where M represents the total number of online and healthy governance modules in the current system. This calculation is performed independently in each module. Since the input data is consistent, the T_avg calculated by each module is theoretically the same, thus forming a unique temperature reference value in the system without a central controller.

[0042] Dynamic adjustment of adaptive virtual impedance: Each module dynamically adjusts its output impedance—i.e., virtual impedance Z_v_i—based on the deviation between its own temperature and the system's average temperature. Z_v_i=Z_v_ref [1+K (T_j_i-T_avg)], Z_v_ref is the system's preset reference virtual impedance, typically designed as inductive (jωL_v), used to set the system's fundamental external characteristics and help suppress high-frequency circulating currents. K is the temperature-impedance adjustment coefficient, a constant greater than zero. This coefficient determines the strength of the temperature feedback, and its value must be determined by a trade-off between dynamic response speed and system stability.

[0043] Closed-loop correction of output instructions: The locally generated raw current command (consisting of i_d_ref and i_q_ref) will be corrected based on the adjusted virtual impedance to form the final command acting on the current loop: Instruction correction formula: i_ref _i=i_ref_original (Z_v_ref / Z_v_i).

[0044] The instruction correction formula is a global formula, which can be expressed as follows in the specific dq coordinate system: i_d_ref _i=i_d_ref_original (Z_v_ref / Z_v_i), i_q_ref _i=i_q_ref_original (Z_v_ref / Z_v_i).

[0045] Since \(Z_{v_i}=Z_{v\_ref}\) [1 + K (T_{j\_i}-T_{avg})]\), the instruction correction formula can also be simplified to: .

[0046] For the overheated load module (\(T_{j\_i}>T_{avg}\)), its virtual impedance \(Z_{v_i}\) will be greater than the reference value \(Z_{v\_ref}\). According to Ohm's law, under the same output voltage instruction, the compensation current output by this module will automatically decrease. The decrease in current leads to a reduction in its conduction loss and switching loss, thus curbing the rising trend of its temperature.

[0047] For the lightly loaded module (\(T_{j\_i}<T_{avg}\)), its virtual impedance \(Z_{v_i}\) will be less than the reference value \(Z_{v\_ref}\). Under the same control instruction, this module will automatically bear more output current.

[0048] The above process is a global negative feedback process. Eventually, the temperatures of all online modules will tend to be the same (\(T_{j\_i}\approx T_{avg}\)), and the virtual impedances of each module will also tend to the reference value, thereby achieving autonomous power sharing and balanced distribution of thermal stress.

[0049] This correction, as a feed - forward link, is seamlessly embedded in the double - closed - loop control of each module. It ensures that the change in virtual impedance can be instantaneously and linearly reflected in the output of the module without affecting the dynamic response performance of the current inner loop.

[0050] Step 3: Generate voltage instructions \(u_{d\_ref}\) and \(u_{q\_ref}\) based on the inner - loop control instruction. Compare the final current instructions \(i_{d\_ref}\) and \(i_{q\_ref}\) obtained in Step 2 with the actual output current of the inverter. The error is calculated by a PI controller and added with a feed - forward decoupling term to generate the voltage instructions \(u_{d\_ref}\) and \(u_{q\_ref}\) output by the controller: , where is the gain coefficient of the current loop, which is \((Kp_i + Ki_i / s)\), \(\omega\) is the fundamental angular frequency of the power grid, 、 are the inner - loop control instructions, \(L_f\) is the value of the output filter inductor of the inverter; \(-\omega\) L_f i_q and \(+\omega\) L_f i_d are the traditional feedback decoupling terms used to cancel the coupling voltages \(+\omega\) L_f i_{q\_ref} and \(-\omega\) L_f i_d_ref is a feedforward decoupling term that can provide compensation voltage in advance according to the instruction, which greatly improves the instruction tracking speed and dynamic performance, and achieves a fast response with almost no overshoot.

[0051] Step 4: Adaptive Resonance Adjustment Hybrid Modulation Harmonic analysis: The local controller performs FFT analysis on the harmonic current components i_d_harm and i_q_harm in step one in real time (using DSP library functions, FFT number 1024), identifies the dominant harmonic order (such as 5th, 7th, 11th) and its amplitude, and obtains the harmonic spectrum.

[0052] Modulation strategy decision: In the fundamental and low harmonic frequency bands, SVPWM is used to maintain high DC voltage utilization and low output ripple; near the identified dominant harmonic frequency band, a specific number of resonant controllers (PR) are dynamically injected, and the modulation mode of that frequency band is switched to DPWM with the optimal switching frequency or specific harmonic elimination PWM (SHEPWM) to specifically improve the gain of that frequency band and optimize switching losses.

[0053] Carrier frequency adjustment: The carrier frequency is dynamically adjusted within a certain range based on the heat sink temperature T_j. When the temperature is high, the carrier frequency is appropriately reduced to decrease switching losses; when the temperature is low, the carrier frequency is increased to improve the output waveform quality.

[0054] Power Output: The modulated PWM wave drives the power switching transistors via optical fiber. The power modules controlling the improved ANPC topology generate the required compensation current, which is ultimately injected into the grid to complete power quality management.

[0055] Those skilled in the art can adjust parameters (such as sampling frequency, MAF window, weighting coefficient k, etc.) according to the above embodiments to adapt to different power grid conditions. This invention achieves high-precision and high-reliability power quality management through dynamic delay compensation, collaborative decision-making, and adaptive modulation.

[0056] The above description is merely the basic principle and preferred embodiment of the present invention. Improvements and substitutions made by those skilled in the art based on the present invention are within the scope of protection of the present invention.

Claims

1. A method for improving power quality in a regional power grid with multiple types of entities connected, characterized in that: Includes the following steps: S01. Sample the three-phase grid voltage and three-phase grid current. Transform the sampled data to a two-phase rotating dq coordinate system using Clark and Park transformations. After Clark transformation, the current in the αβ coordinate system is i_α and i_β. After Park transformation, the direct-axis component of the voltage in the dq coordinate system is U_d and the quadrature-axis component is U_q. The direct-axis component of the current is i_d and the quadrature-axis component is i_q. After performing phase lead compensation on i_α and i_β, Park transformation is performed to obtain the fundamental current components i_d_fund and i_q_fund without delay. S02. Calculate the effective voltage deviation ΔU and the total harmonic distortion (THD) of the current. The global mode command, along with ΔU and THD, is input to the local decision-maker. The local decision-maker dynamically generates the harmonic compensation weighting coefficient K_h and the fundamental reactive current command i_q_svg_ref based on the preset priority logic. When the global mode instruction is voltage regulation priority mode and |ΔU| is greater than the set threshold, the local decision-maker increases i_q_svg_ref and decreases K_h, so that the governance module can give priority to the SVG function to support the voltage. When the global mode instruction is harmonic suppression priority mode and THD is greater than the set threshold, the local decision-maker increases K_h and decreases i_q_svg_ref, so that the governance module can give priority to the APF function to suppress harmonics. S03. The governance module adopts dual closed-loop control, with an outer loop being a voltage loop and an inner loop being a current loop. The voltage outer loop contains two independently operating outer loop controllers: a DC voltage outer loop and an AC voltage outer loop. The DC voltage outer loop outputs the fundamental active current d-axis component i_d_dc_ref; the AC voltage outer loop outputs the fundamental reactive current q-axis component i_q_fund_ref. i_d_dc_ref and i_q_fund_ref are synthesized with harmonic commands and then corrected by thermal equalization to obtain the final inner loop commands i_d_ref and i_qref. S04. Generate voltage commands u_d_ref and u_q_ref based on the inner loop command: ， in ω is the gain coefficient of the current loop, L_f is the fundamental angular frequency of the power grid, and L_f is the output filter inductance value of the inverter. S05. Perform Park and Clark inverse transformations on the voltage command to obtain a three-phase modulation wave. Combine this with a dynamic modulation strategy to generate a PWM signal. Drive the converter to output compensation current based on the PWM signal.

2. The regional power quality management method for multi-type entity access as described in claim 1, characterized in that: The thermal balance correction in step S03 is as follows: all governance modules communicate with each other, sending data frames including heatsink temperatures, calculating the average temperature of all governance modules, and obtaining the final inner loop instructions i_d_ref and i_q_ref based on the average temperature. , This is the final inner loop instruction for the i-th current loop. The inner loop instruction before correction is defined as follows: T_j_i is the heatsink temperature of the i-th governance module, T_avg is the average heatsink temperature of all governance modules, and K is the temperature-impedance adjustment coefficient.

3. The regional power quality management method for multi-type entity access as described in claim 1, characterized in that: The local decision-maker dynamically generates the harmonic compensation weighting coefficient K_h and the fundamental reactive current command i_q_svg_ref based on the global mode command, ΔU, and THD. K_h is used to generate the harmonic compensation command, and i_q_svg_ref is used to adjust the voltage reference value of the AC voltage outer loop. ， in , For d-axis and q-axis harmonic compensation commands, , These are d-axis and q-axis harmonic currents with no time delay. , U_ref=Unom+Ku iq_svg_ref, Wherein, U_ref is the voltage reference value of the AC voltage outer loop, that is, the voltage target that the PCC point is expected to maintain; Unom is the rated voltage, that is, the nominal voltage of the power grid during normal operation; iq_svg_ref is the fundamental reactive current command, reflecting the reactive demand intensity under the current mode; and Ku is the voltage-reactive conversion coefficient, which is the proportional factor that converts the reactive current target into the voltage offset.

4. The regional power quality management method for multi-type entity access as described in claim 1, characterized in that: In step S01, the phase delay τ_total for the lead phase compensation is: , The formula for leading phase compensation is: ， Where N is the moving average filter window length, k is the sampling time, f_s is the sampling frequency, T_s is the control period; i_α_compensated(k) and i_β_compensated(k) are the α-axis currents after phase compensation, ω is the grid current angular frequency, and i_α(k) and i_β(k) are the α-axis currents before phase compensation.

5. The regional power quality management method for multi-type entity access as described in claim 1, characterized in that: In step S02, the voltage RMS deviation ΔU is calculated as follows: the fundamental voltage RMS value is calculated based on the voltage direct-axis component u_d in the dq coordinate system. And thus obtain U nom This is the rated voltage.

6. The regional power quality management method for multi-type entity access as described in claim 1, characterized in that: The total harmonic distortion (THD) of current is calculated as follows: based on the fundamental current components i_d_fund and i_q_fund without delay, the harmonic current components i_d_harm and i_q_harm are calculated. The fundamental current component and harmonic current component without delay are converted into the total effective value of the fundamental current I_fund and the total effective value of the harmonic current I_harm, and then calculated. .

7. The regional power quality management method for multi-type entity access as described in claim 3, characterized in that: The AC voltage outer loop control process is as follows: The three-phase voltage at the PCC point is sampled in real time. After phase-locked loop (PLL) and dq transformation, the d-axis voltage component u_d is obtained. The deviation ΔU between u_d and U_ref is calculated. The deviation ΔU is fed into the outer loop, and the outer loop output i_q_fund_ref is: ， Where K_droop is the droop coefficient, Q is the reactive power output of the machine, Q_avg is the average reactive power of the system, and Kp and Ki are the AC proportional gain and AC integral gain of the outer loop. The outer loop output and harmonic compensation command are combined to form the initial base value i_q_ref_original for the inner loop: i_q_ref_original=i_q_fund_ref+i_q_h_ref.

8. The regional power quality management method for multi-type entity access as described in claim 3, characterized in that: The DC voltage outer loop control process is as follows: The DC bus voltage U_dc is sampled in real time. The sampled value is compared with the given voltage reference value U_dc_ref to obtain the error e_dc(t) = U_dc_ref - U_dc(t). This error is then fed into the outer loop, and the outer loop output i_d_dc_ref(t) is: ， Where Kp_d and Ki_dc are the DC proportional gain and DC integral gain of the outer loop, respectively. The outer loop output and harmonic compensation command are combined to form the initial base value i_d_ref_original for the inner loop i_q_ref: i_d_ref_original=i_d_dc_ref+i_d_h_ref.

9. The regional power quality management method for multi-type entity access as described in claim 1, characterized in that: Combining a dynamic modulation strategy, the PWM signal is generated as follows: after obtaining the time-delayed fundamental current components i_d_fund and i_q_fund, according to... Harmonic current components are acquired, and FFT analysis is performed on them to identify the dominant harmonic order and its amplitude, thus obtaining the harmonic spectrum. SVPWM modulation is applied to the fundamental and lower harmonic frequency bands. A resonant controller of a specific order is dynamically injected near the identified dominant harmonic frequency band, and the modulation method for this frequency band is switched to DPWM with the optimal switching frequency or specific harmonic elimination PWM. The carrier frequency is dynamically adjusted within the specified harmonic spectrum range based on the heat sink temperature. A PWM signal is generated by combining the voltage signal from step S04 and the modulation method. The PWM signal enters the converter, controlling the power module of the improved ANPC topology to generate the required compensation current, which is finally injected into the power grid to complete power quality management.

10. The regional power quality management method for multi-type entity access according to claim 9, characterized in that: Within the specified carrier frequency range, if the radiator temperature is higher than the set value, the carrier frequency is reduced; if the radiator temperature is lower than the set value, the carrier frequency is increased.

Images