Power harmonic treatment method for improving active power filter by novel multi-level converterTechnical Field
The invention relates to the technical field of Power harmonic treatment, in particular to the technical field of Active Power Filter (APF) treatment of Power harmonics, and discloses a Power harmonic treatment method for improving an Active Power Filter by using a novel Multilevel Converter (MMC). On the premise of not changing a harmonic detection technology, the voltage grade and the current tracking capability of the active power filter are improved by adopting a modular multilevel converter topology and a loop suppression strategy and optimizing a current tracking module, so that the active power filter can be applied to higher-voltage occasions, the flexibility of the active power filter is improved, and unnecessary personnel and economic losses caused by compensation failure due to damage of a bridge arm of a traditional bridge circuit are reduced.
Background
With the popularization and application of modern power electronic technology, a large amount of harmonic waves are generated, the power supply quality is seriously influenced, the working performance and the service life of power generation equipment and power utilization equipment are reduced, and even the safety of a power system is endangered. Thus, power harmonic governance has become a hot issue in terms of power quality. The active power filter has the advantages of tracking and compensating harmonic waves with variable frequency and amplitude, compensating each harmonic wave, inhibiting flicker, compensating reactive power and preventing the filter characteristic from being influenced by system impedance, and is widely concerned in the field of power harmonic wave treatment at present.
However, due to the limitation of power electronic devices, active power filters are only applied to lower voltage levels at present, and are mainly connected to a high-voltage system through a transformer in high-voltage and high-capacity occasions, and then form high capacity in a high-current mode, but the transformer link has adverse effects on the compensation of the power quality, and particularly higher harmonic compensation is difficult. In recent years, people have tried to improve the voltage class and capacity of an active power filter by various methods, such as cascading the active power filter, improving a topological structure, and the like, but the problems that the voltage class is not changed greatly due to structural limitation, harmonics generated by the topological structure are difficult to treat, and the like still exist, and technical innovation is urgent.
The modularized multi-level converter as a newly generated technology in recent years has the advantages of high modularization, easiness in expansion and convenience in improving the voltage grade, and has obvious application advantages in the fields of medium-high voltage and large capacity. Therefore, a Multilevel Converter-Active Power Filter (MMC-APF) method is becoming a potential research direction of the Active Power Filter in high-voltage occasions.
However, although the MMC-APF can be applied to high-voltage and large-capacity occasions, the compensation current has errors due to the problems of poor tracking speed and low precision of the traditional current tracking method, and meanwhile, the MMC-APF generates circulating current due to the special structure of the MMC-APF, so that the size of the compensation current is also influenced, and the compensation performance is influenced. Therefore, the APF with high voltage, large capacity and excellent compensation performance is applied to the power harmonic suppression.
On the basis of fully analyzing the technical characteristics of the existing APF, the invention provides a power harmonic treatment method for improving an active power filter by using a novel multi-level converter as shown in figure 1 by combining the fact that the APF cannot obtain a good application effect in high-voltage and large-capacity occasions, so that the APF is applied to medium-high-voltage and large-capacity occasions, the fault tolerance of the APF is improved, and the accurate harmonic treatment can be realized.
Disclosure of Invention
The technical problem solved by the invention is as follows: in view of the defects in the prior art, the advantages of the MMC are combined, the purpose of the invention is to provide a power harmonic treatment method for improving an active power filter by using a novel multi-level converter aiming at the situation that the APF can only be applied to the occasions with low voltage level, medium and small capacity and in order to realize the application of the APF to the occasions with high voltage and large capacity and ensure the compensation effect and flexibility of the APF. The method is characterized in that a multi-level converter is used for improving The construction of a main circuit model of an active power filter, a circulation current restraining link based on second-order generalized integral (SOGI) and a fuzzy Proportional Integral (PI) -repetitive control optimization current tracking control algorithm are designed on The premise of not changing The original harmonic detection technology, so that The requirements of high voltage and large capacity are met, The current tracking speed is increased, The flexibility is increased, and The power quality is improved.
The technical scheme for solving the problems is as follows: a power harmonic treatment method for improving an active power filter by using a novel multi-level converter is characterized by comprising the following steps:
step 1, constructing a main circuit model of an improved active power filter by using a multilevel converter:
1) creating an H half-bridge submodule;
2) the bridge arm of the original APF three-phase bridge circuit is replaced by a structure in which a plurality of H half-bridge submodules and inductors are connected in series, the flexible adjustment of the number of the H half-bridge submodules is realized, and the H half-bridge submodules of the bridge arm can be replaced with each other according to the fault condition, so that the voltage grade and the capacity of the APF are increased, and the fault tolerance of the APF is improved;
3) when the voltage grade and the capacity of the APF are increased, the output end of the upper bridge arm and the output end of the lower bridge arm of each phase are connected with a direct current capacitor, so that the working stability of the APF is improved;
the power supply of the MMC-APF main circuit is three-phase voltage sources Vsa, Vsb and Vsc with the same amplitude and phase difference of 120 degrees, one end of each phase of voltage source is connected with a neutral point, the other end of each phase of voltage source is respectively connected with one end of an MMC-APF input reactor, the other end of each MMC-APF input reactor is respectively connected with the input ends of an upper bridge arm and a lower bridge arm of each phase of voltage source, the output ends of the upper bridge arm and the lower bridge arm of each phase of voltage source are respectively connected with two ends of a direct current capacitor, each phase of voltage source comprises an upper bridge arm and a lower bridge arm, and each bridge arm is formed by connecting n submodules in series; SMn represents each submodule of an upper bridge arm and a lower bridge arm of a three-phase voltage source, and n is 1-z to represent an nth submodule;
for the basic composition unit of the modular multilevel converter, the MMC submodule is usually in an H half-bridge structure and is formed by connecting two IGBTs in series and then connecting the two IGBTs in parallel with a capacitor C. When the main circuit runs, bridge arm current flows in two directions between the two IGBTs, and a required modulation waveform is fitted through switching cooperation of the sub-modules of the upper bridge arm and the lower bridge arm;
step 2, restraining the circulating current of the multilevel converter based on a second-order generalized integral multiple harmonic filter:
1) extracting a double frequency component in the MMC-APF circulating current by adopting SOGI;
2) obtaining harmonic components by combining negative feedback and low-pass filtering links;
3) a quasi-Proportional Resonance (PR) controller is adopted to improve the multi-harmonic circulating current suppression effect;
step 3, current tracking optimization based on fuzzy proportional integral repetitive control:
1) establishing current inner loop control on the basis of PI control to eliminate the coupling influence;
2) and a repetitive controller and a fuzzy controller are designed, so that the accurate and real-time tracking of the command current is realized, and the effectiveness and the real-time performance of harmonic treatment are improved.
Further, in thestep 1, the construction of the main circuit model of the active power filter improved by the multilevel converter comprises:
1) on the basis of an APF three-phase bridge circuit structure, two Insulated Gate Bipolar Transistors (IGBT) are connected in series and connected in parallel with a capacitor to form an H half-bridge submodule;
2) the bridge arm of the original APF three-phase bridge circuit is replaced by a series structure of a plurality of H half-bridge submodules and an inductor, the number of the H half-bridge submodules is flexibly adjusted, and the H half-bridge submodules of the bridge arm can be replaced with each other according to the fault condition, so that the voltage grade and the capacity of the APF are increased, and the fault tolerance of the APF is improved;
3) when the voltage grade and the capacity of the APF are increased, the output end of the upper bridge arm and the output end of the lower bridge arm of each phase are connected with a direct current capacitor, so that the working stability of the APF is improved;
4) on the basis of the established MMC-APF circuit structure, according to the circuit principle, the MMC model under the dq coordinate system for obtaining the MMC-APF is as follows:
in the formula ud、uqD, q-axis components, e, of the three-phase ac supply voltage, respectivelyd、eqRespectively MMC equivalent internal electromotive force exD, q-axis components of (i)d、iqD-axis components and q-axis components of output current of the APF converter respectively;
the frequency domain mathematical model is:
according to the formula, a mathematical model of the multi-level converter in the frequency domain can be made.
Further, in the step 2, the MMC circulating current suppression based on the SOGI multi-harmonic filter includes:
1) extracting a second harmonic component in MMC-APF circulating current by adopting an SOGI multi-harmonic filter, filtering a high-frequency even-order component by a negative feedback and low-pass filtering link to obtain a direct current quantity, and inputting the direct current quantity into initial circulating current by utilizing the negative feedback link to further obtain a harmonic component;
a multi-harmonic filter (MHF) is composed of an SOGI and an integrating element, and has a transfer function of
In the formula: omegahThe harmonic separation point of the employed MHF; the 1/s link can amplify the direct current component, so that the direct current component can be separated more conveniently; k is a radical ofhIs an adjustment factor; dynamic performance of MHF is then compared with kh(ii) related;
2) filtering out frequency doubling and frequency quadrupling harmonics by using a quasi-PR controller; having a transfer function of
Wherein: k is a radical ofpIs a proportionality coefficient; k is a radical ofrIs the resonance coefficient; omegarIs the resonant frequency; omegar=kω0;ωcIs the cut-off frequency; s is a complex frequency domain variable.
When ω is ω ═ ωrWhile its amplitude is kp+krThe phase angular displacement is 0, so that the non-static tracking of specific frequency can be realized; taking factor of gain and bandwidth into consideration, take omegac=π;
Because the circulation is mainly the second harmonic and the fourth harmonic, the invention mainly controls the second harmonic and the fourth harmonic, and adopts a circulation restraining quasi-PR controller of MHF to obtain omegar=2ω0And ωr=4ω0。
Further, in thestep 3, the current tracking optimization based on the fuzzy PI-repetitive control comprises:
1) establishing current inner loop control, and introducing a PI transfer function of the inner loop control as follows:
wherein KPR, L is equivalent resistance and equivalent inductance which are proportional parameters of the PI controller; kIIs an integral parameter of the PI controller; l is 14mH, and R is 0.08 Ω.
2) Design repetitive controller and fuzzy controller, realize accurate, the real-time tracking of instruction current, promote the validity and the real-time nature of harmonic treatment:
repeating controller cycle delay link z-NA repeated internal mold and a corrector S (z);
in the active filter, the power frequency of the power grid is f-50 Hz, and the switching frequency is fs=25600Hz。
For a repeating internal model, its transfer function can be expressed as:
wherein q (z) is an attenuation filter, and q (z) is 0.95; z is a radical of-NFor delay compensation.
In the repetitive control, a corrector S (z) only needs to design a second-order low-pass filter to carry out amplitude attenuation on the PI inner ring, and high-frequency disturbance of the system is suppressed; the transfer function of the second order low pass filter is:
in the formula of omeganThe cut-off frequency is usually chosen to be half the frequency of the highest order waves of the system, and ω is the highest frequency required to compensate for the 50 th harmonicnTaking 2500 pi rad/s; xi is a damping coefficient, which is 0.707; further, the discrete domain transfer function of f (z) is obtained as:
meanwhile, 3-beat lead compensation z is introduced to correct PI inner loop and after the phase generated by Low Pass Filter (LPF) is full3And combining system stability to repeatedly control gain KrSet to 0.98, therefore, s (z) is expressed as:
S(z)=0.98z3F(z)
secondly, according to the requirements in the process, the dynamic response speed is high, the dynamic response speed caused by a delay link is reduced, a fuzzy rule is established, 7 membership function big (NB), Negative Middle (NM), Negative Small (NS), Zero (ZO), Positive Small (PS), Positive Middle (PM) and positive large (PB) are added in the range, different fuzzy languages are given to the 7 membership function according to the deviation value, and the designed delta KP、ΔKIThe control rule of (2) is shown in table 1, and the membership function of the fuzzy control adopts a trigonometric function;
after fuzzy reasoning, the output value is defuzzified and converted into an accurate value, and finally, delta K is outputP、ΔKI(ii) a Parameter KP、KIThe calculation formula of (a) is as follows:
the invention has the beneficial effects that: by means of the technical scheme, the novel multi-level converter improves the power harmonic control method of the active power filter, and can realize that the APF is applied to high-voltage and large-capacity occasions without changing the original harmonic detection method, the compensation current is accurately tracked, the accurate and reliable harmonic compensation capability of the system is ensured, and the fault tolerance of the APF is improved.
Drawings
FIG. 1 is a schematic diagram of the power harmonic suppression method of the present invention using a novel multi-level converter to improve an active power filter;
FIG. 2 is a schematic diagram of the MMC-APF main circuit of the present invention;
FIG. 3 is a schematic diagram of an H-half bridge submodule structure according to the present invention;
FIG. 4 is an overall block diagram of MMC circulating current suppression based on SOGI multi-harmonic filter;
FIG. 5 is a schematic diagram of current inner loop control;
FIG. 6 is a PI control schematic;
FIG. 7 is a schematic diagram of repetitive control;
FIG. 8 is a fuzzy PI-repeat control block diagram;
FIG. 9 is a compensation current waveform display of MMC-APF filtering effect based on fuzzy PI-repetitive control;
FIG. 10 is a load current demonstration of MMC-APF filtering effect based on fuzzy PI-repetitive control;
FIG. 11 is a harmonic current waveform display of MMC-APF filtering effect based on fuzzy PI-repetitive control;
FIG. 12 is a compensation current waveform distortion rate display based on MMC-APF filtering effect of fuzzy PI-repetitive control;
table 1 is a fuzzy control rule table.
In the figure: 1 MHF, 2 quasi PR controllers, 3 current inner rings, 4 MMC mathematical models, 5 repetitive controllers and 6 fuzzy PI controllers.
Detailed Description
The present invention will be described in detail below with reference to the drawings and examples.
Referring to fig. 1 to 12 and table 1, inembodiment 1, the method for implementing power harmonic suppression by using a novel multilevel converter to improve an active power filter includes the following steps:
step 1, constructing a main circuit model of an improved active power filter by using a multilevel converter:
1) creating an H half-bridge submodule;
on the basis of an APF three-phase bridge circuit structure, two Insulated Gate Bipolar Transistors (IGBT) (T1 and T2 in the figure 3) are connected in series and connected with a capacitor C in parallel to form an H half-bridge submodule;
2) the bridge arm of the original APF three-phase bridge circuit is replaced by a series structure of a plurality of H half-bridge submodules and an inductor, the number of the H half-bridge submodules is flexibly adjusted, and the H half-bridge submodules of the bridge arm can be replaced with each other according to the fault condition, so that the voltage grade and the capacity of the APF are increased, and the fault tolerance of the APF is improved;
3) when the voltage grade and the capacity of the APF are increased, the output end of the upper bridge arm and the output end of the lower bridge arm of each phase are connected with a direct current capacitor, so that the working stability of the APF is improved;
4) for the basic composition unit of the modular multilevel converter, the MMC sub-module usually adopts an H half-bridge sub-module, and when the main circuit runs, the bridge arm current iarmAnd the two IGBTs flow in two directions, and the required modulation waveform is fitted through the switching matching of each H half-bridge submodule of the upper bridge arm and the lower bridge arm. According to the different switch states of T1 and T2, the H half bridge sub-module can be roughly divided into three operation modes of 'on', 'off' and 'bypass', and the bypass state only appears when the system starts to charge the capacitor or is in an abnormal state such as a fault.
As shown in fig. 2-3, a power supply of the MMC-APF main circuit is three-phase voltage sources Vsa, Vsb and Vsc with the same amplitude and a phase difference of 120 °, one end of each phase voltage source is connected to a neutral point, the other end of each phase voltage source is connected to an end of an MMC-APF input reactor, the other end of each MMC-APF input reactor is connected to input ends of an upper bridge arm and a lower bridge arm of each phase voltage source, output ends of the upper bridge arm and the lower bridge arm of each phase voltage source are connected to two ends of a direct current capacitor, each phase voltage source comprises an upper bridge arm and a lower bridge arm, and each bridge arm is formed by connecting n H half-bridge submodules in series; namely SMnThe power supply comprises an upper bridge arm and a lower bridge arm, wherein the upper bridge arm and the lower bridge arm represent an upper bridge sub-module and a lower bridge sub-module of a three-phase voltage source, N is 1-z represents an nth H half bridge sub-module, Ia, Ib and Ic represent output currents of the three-phase voltage source Vsa, Vsb and Vsc respectively, Ls is an MMC-APF input reactor, a neutral point of the three-phase voltage source is recorded as N, and a direct-current capacitor is recorded as C2,C3。
MMC-APF mathematical model:
defining an H half-bridge submodule switch state function K:
the H half-bridge sub-module output voltage and dc side current can be expressed as:
wherein u isdc-xyzH half-bridge submodule dc-side capacitor voltage (z ═ 1, 2, 3,. n), u, for MMCxyz,ixyzRespectively an output voltage at the AC side and a current at the DC side of the H half-bridge submodulexy(y represents upper and lower arms) is the arm current.
And because the bridge arm voltages and the output voltage of the H half-bridge submodule have the following relationship:
wherein u isxyIs bridge arm voltage uxyzAnd outputting voltage for the alternating current side of the H half-bridge submodule.
According to the output characteristics of each bridge arm, the bridge arms can be equivalent to controllable voltage sources. Bridge arm inductance L0Having an internal resistance of R0Taking phase a as an example, analyzing a loop formed by an upper bridge arm, a lower bridge arm and a virtual direct current common bus by using kirchhoff voltage law:
Udcis a DC bus voltage uap,uanA phase upper and lower bridge arm voltages, iap,ianThe current of the upper bridge arm and the lower bridge arm of the phase a.
Making a difference between the two formulas
The network side alternating current power supply and the output of the MMC current converter have the following relations:
in the formula usaFor the grid side a phase voltage, icaIs a side of a netPhase a current, ucaTo output voltage for the inverter.
The formula (5) may be substituted for the formula (6):
according to kirchhoff's current law, the following relations exist between the currents of the upper bridge arm and the lower bridge arm and the output current of the current converter:
ica=ian-iap (8)
the formula (6) may be substituted for the formula (7):
finally, let
The formula (9) can be collated:
because MMC transverter has qualified symmetry in the structure, consequently can deduce its mathematical model under three-phase static coordinate system to be:
however, the MMC inverter has more variables in the three-phase stationary coordinate system, and the voltage and current are ac time-varying components, so as to reduce the analysis difficulty, equation (12) is converted into the two-phase rotating coordinate system. Meanwhile, in order to make the vector module values before and after conversion the same, introducing a constant amplitude Park transformation conversion matrix as shown in formula (13):
the MMC mathematical model 4 under dq coordinate system is:
in the formula ud、uqD, q-axis components, e, of the three-phase ac supply voltage, respectivelyd、eqRespectively MMC equivalent internal electromotive force exD, q-axis components of (i)d、iqThe components of the output current of the MMC converter are the d-axis component and the q-axis component respectively. To facilitate the design of MMC-APF system controller, Laplace transform is used to transform equation (14) into frequency domain
A mathematical model of the MMC converter in the frequency domain as shown in fig. 5 can be made according to equation (15).
In the step 2, MMC circulation suppression of the SOGI-based MHF1 comprises the following steps:
1) circulation is generated due to the structural particularity of the MMC-APF, an MMC circulation restraining method based on the SOGI is adopted, and the structure of the MMC circulation restraining method is shown in figure 4. The SOGI-based MMC loop suppression consists of two parts, namely an SOGI-based MHF1 and a quasi-PR controller 2. The MHF1 firstly adopts SOGI to extract a double frequency component, compares the double frequency component with the initial circulating current through a negative feedback link, then filters the high frequency even-order component through a low-pass filtering link to obtain a direct current quantity, and inputs the direct current quantity into the initial circulating current through the negative feedback link to obtain a harmonic component.
MHF1 is composed of an SOGI and an integration element, and has a transfer function of
In the formula: omegahThe harmonic separation point for the MHF1 employed; the 1/s link can amplify the direct current component to make it more squareIs separated out conveniently; k is a radical ofhIs an adjustment factor; dynamic performance of MHF1 is then compared with khIt is related.
2) Filtering out frequency doubling and frequency quadrupling harmonics by using a quasi-PR controller 2;
in order to improve the stability of the system, eliminate the problems that the ideal PR bandwidth is too narrow and the amplification at a resonant frequency point is small, and effectively and accurately eliminate harmonic waves, the quasi-PR controller 2 is utilized to realize the circulation current suppression. Having a transfer function of
Wherein: k is a radical ofpIs a proportionality coefficient; k is a radical ofrIs the resonance coefficient; omegarIs the resonant frequency; omegar=kω0;ωcIs the cut-off frequency; s is a complex frequency domain variable.
When ω is ω ═ ωrWhile its amplitude is kp+krAnd the phase angular displacement is 0, so that the non-static tracking of a specific frequency can be realized. Taking factor of gain and bandwidth into consideration, take omegac=π。
Because the circulation is mainly the second harmonic and the fourth harmonic, the main control targets of the invention are the second harmonic and the fourth harmonic, and the circulation suppression quasi-PR controller 2 adopting MHF1 takes omegar=2ω0And ωr=4ω0。
In thestep 3, the current tracking optimization based on the fuzzy PI-repetitive control comprises the following steps:
1) establishing current inner loop control, introducing PI transfer function of inner loop control
As the MMC-APF mathematical model introduces the cross-coupling term omega Li in the process of converting the abc coordinate system into the dq coordinate systemdAnd ω LiqAnd on the other hand, a feedforward cross decoupling method is adopted to establish current inner loop control and eliminate the coupling influence.
On the d-axis channel, it can be seen that ω Li is a state quantity due to the feedforwarddThe coupling terms in the MMC mathematical model 4 are offset, and finally the independent control of the d axis and the q axis is realized. In view of the fact that the object under control is a first-order inertial system,an PI control strategy with simple use principle and strong operability is tried, and a currentinner ring 3 transfer function based on PI control can be obtained as
Wherein KPR, L is equivalent resistance and equivalent inductance which are proportional parameters of the PI controller; kIIntegrating parameters for the PI controller; in this specification, L is 14mH, and R is 0.08 Ω.
2) Design repetitive controller and fuzzy controller, realize accurate, the real-time tracking of instruction current, promote the validity and the real-time nature of harmonic treatment:
the repeated control improves the accuracy of tracking steady-state error and the stability of motion of the system, and is added on the basis of PI control. The principle of repetitive control is shown in FIG. 7, where therepetitive controller 5 is formed by a cycle delay element z-NThe repeated internal model and the corrector S (z). Wherein z is-NIs the action that the system does after delaying for a period of one cycle, namely: the error information detected in the previous period is to generate control influence in the previous period, wherein N is the total number of sampling points in the period, the power frequency of a power grid in the active filter is f-50 Hz, and the switching frequency is fs=25600Hz。
For a repeating internal model, the transfer function can be expressed as:
where q (z) is an attenuation filter that can be set as either a low pass filter or a constant less than 1, assuming q (z) takes on the real number x, equation (20) is expressed as a difference equation:
u(k)=e(k)+x*u(k-N) (21)
according to the formula (21), the error e (k) is integrated by repeating the internal model by taking the power frequency period as the step length until the error is less than (1-x) times u (k), and the margin is reserved for the integration link due to the setting of Q (z), so that the interference caused by high-frequency gain can be inhibited. The repeated control has high gain and zero phase shift characteristics at the power frequency and frequency multiplication positions thereof, namely the repeated internal model is stable: however, as q (z) increases, the system control gain also increases gradually, so that part of the stability is sacrificed while improving the steady-state accuracy, and q (z) is 0.95 in comprehensive consideration.
The function of the corrector S (z) in the repetitive control is to ensure that the controlled object has zero gain and zero phase shift in the middle and low frequency bands, and simultaneously enhance the high-frequency attenuation characteristic of the forward channel. Because the controlled object has good frequency response characteristic in a low frequency band, only a second-order low-pass filter needs to be designed to perform amplitude attenuation on the PI inner ring, and high-frequency disturbance of the system is suppressed. The transfer function of the second order low pass filter is:
in the formula of omeganFor the cut-off frequency, the value of it will affect the gain of the low frequency band, and is generally selected according to half of the highest harmonic frequency of the system, and if the highest requirement is to compensate 50 harmonics, then ω isnTake 2500 π rad/s. ξ is the damping coefficient, which determines the rate of rise of the signal, here taken to be 0.707. Further, the discrete domain transfer function of f (z) is obtained as:
meanwhile, 3-beat lead compensation z is introduced for correcting the phase produced by the PI inner loop and the LPF after full phase3And combining system stability to repeatedly control gain KrSet to 0.98, therefore, s (z) can be expressed as:
S(z)=0.98z3F(z) (24)
and finally, on the basis of PI-repetitive control, fuzzy and PI compound control is introduced, the dynamic response speed of thefuzzy PI controller 6 is high, the PI parameter is adjusted in real time, and the phenomenon that the dynamic response speed is slowed down due to a delay link is reduced. Therefore, the compensation current is accurately tracked in real time.
KPThe value of (A) directly affects the response speed of the system, KPThe larger the value is, the faster the system adjustment speed is, and the shorter the response time is, but the excessive value causes the overshoot and oscillation frequency of the system to increase, so that the response time is prolonged, and meanwhile, the system is too sensitive, and the dynamic performance is reduced; increase of KIThe value can effectively reduce steady-state error and enhance the dynamic stability of the system, but the system adjusting time is increased due to the overlarge value. When the error and the error change rate are considered together, e and ecIn the same number, K is increasedP(ii) a Conversely, decrease KP. When K isIToo large, it may cause an increase in system overshoot. To prevent integral saturation, let K if the error e is largeIIs equal to zero; if e is smaller, let KIAnd act to reduce the systematic error by increasing it as e decreases.
Secondly, according to the requirements in the process, to realize high dynamic response speed and reduce low dynamic response speed caused by a delay link, 7 membership function big (NB), Negative Middle (NM), Negative Small (NS), Zero (ZO), Positive Small (PS), Positive Middle (PM) and Positive Big (PB) are added in the range, different fuzzy languages are given to the 7 membership function according to the size of the deviation value, and the designed delta K is providedP、ΔKIThe control rule of (2) is shown in table 1, and the membership function of the fuzzy control is a trigonometric function.
After fuzzy reasoning, the output value is defuzzified and converted into an accurate value, and finally, delta K is outputP、ΔKI. Parameter KP、KIThe calculation formula of (a) is as follows:
to verify the feasibility of the above theory, simulation experiments were performed under Matlab. In fig. 9-12, the waveforms are the compensation current waveform, the load current, the harmonic current waveform and the compensation current waveform distortion rate in sequence, and it can be seen that the load current has obvious distortion.
TABLE 1