Pulse load compensation device and method based on fuzzy predictive control

Abstract

The invention discloses a pulse load compensation device and method based on fuzzy predictive control, and relates to the technical field of pulse load power compensation. In order to solve the technical problems that in the prior art, an existing load compensation control strategy is too large in size and weight of an output capacitor, the stability and the electric energy quality of a power supply system are affected, and the control response speed is low, the technical scheme provided by the invention is that a pulse load compensation device based on fuzzy prediction control is provided. Comprising a Y-type rectifier for converting three-phase alternating current into direct current; the output capacitor is used for absorbing current disturbance and maintaining voltage stability under a common load; the DC-DC converter is used for realizing power decoupling when the pulse load is connected; the energy storage capacitor provides a current difference when the pulse load is connected to the direct current end; and the control module controls the DC-DC converter, tracks the reference value of the pulse current in real time, realizes current tracking, and reduces the impact of the pulse load on the direct current side voltage of the system. The method is suitable for the work of pulse load power compensation.

Description

Technical Field

[0001] The invention relates to the technical field of pulse load power compensation, and in particular to a pulse load compensation strategy based on fuzzy predictive control. Background Art

[0002] In recent years, with the widespread application of information equipment and the continuous advancement of power electronics technology, the research on pulse loads has gradually become a hot topic. Pulse loads are a type of load with rapid and sudden changes in instantaneous power. They have periodic and low-frequency characteristics and are widely used in military equipment such as phased array radars, lasers, and electromagnetic weapons. When this type of load is running, its instantaneous power changes greatly, which often leads to power quality problems, such as DC side voltage fluctuations and three-phase current distortion.

[0003] At present, the research on pulse loads mainly focuses on two aspects:

[0004] 1. Simulation and experimental study of pulse load operating characteristics: By establishing a mathematical model and conducting simulations, the current and voltage waveforms of pulse loads under different working conditions and their impact on the power supply system are analyzed. The changes in system stability after the pulse load is connected are verified experimentally, and its impact on the power quality of the power supply system is studied.

[0005] 2. Research on compensation control strategy: In order to reduce the impact of pulse load on the DC side voltage of the system, the existing technology mostly uses output capacitors to absorb current pulsation. However, due to the drastic change in the instantaneous current of the pulse load, the output capacitor needs to have sufficient capacity to effectively compensate, resulting in its large size and weight, which increases the complexity of the system. At the same time, the traditional proportional integral (PI) control method is difficult to cope with the rapid changes of pulse loads, and the accuracy and response speed of current tracking are limited.

[0006] Although the existing compensation control strategy has a certain improvement effect on pulse loads, the following problems still exist:

[0007] 1. The volume and weight of the output capacitor are too large: In order to compensate for the current fluctuations caused by the pulse load, the output capacitor has a high capacity requirement, which leads to its large volume and weight, limiting the compactness and flexibility of the system.

[0008] 2. Power quality issues: The access of pulse loads causes significant voltage fluctuations on the DC side and distortion of the three-phase AC current, affecting the stability and power quality of the power supply system.

[0009] 3. Slow control response speed: Traditional PI control is slow to respond to the rapid start and stop of pulse loads and cannot track current changes in time, affecting the decoupling effect and voltage stability of the system. Summary of the invention

[0010] In order to solve the technical problems existing in the prior art, the existing load compensation control strategy has the problem that the volume and weight of the output capacitor are too large, affecting the stability and power quality of the power supply system and the control response speed is slow, the technical solution provided by the present invention is:

[0011] A pulse load compensation device based on fuzzy predictive control, the device comprising:

[0012] Three-phase AC power supply for providing electrical energy;

[0013] Y-type rectifier, used to convert three-phase AC power into DC power;

[0014] Output capacitor, used to absorb current disturbances and maintain voltage stability under normal load;

[0015] DC-DC converter, used to achieve power decoupling when pulse load is connected;

[0016] Energy storage capacitor, used to provide the current difference when the pulse load is connected to the DC terminal;

[0017] The control module is used to control the DC-DC converter, track the reference value of the pulse current in real time, realize current tracking, and reduce the impact of the pulse load on the DC side voltage of the system.

[0018] Furthermore, the input end of the Y-type rectifier is connected to the three-phase AC power supply, and the output end is connected to the DC output end.

[0019] Furthermore, the output capacitor is connected in parallel with the DC output terminal.

[0020] Furthermore, the DC-DC converter is connected in parallel with the DC output terminal.

[0021] According to the described pulse load compensation device based on fuzzy predictive control, the DC-DC converter is implemented by a buck-boost bidirectional converter.

[0022] Furthermore, the energy storage capacitor is connected in parallel with the DC-DC converter.

[0023] Furthermore, the control module tracks the reference value of the pulse current in real time through the model prediction controller and the fuzzy controller; and realizes the tracking of the periodic changes of the current and the energy storage capacitor voltage through the adaptive online adjustment of the weight coefficient.

[0024] A pulse load compensation method based on fuzzy predictive control, the method is implemented based on the device, comprising:

[0025] Steps to convert AC to DC;

[0026] The steps to absorb the current disturbance of DC power and maintain the voltage stability under normal load;

[0027] The step of decoupling the power when the pulse load is connected;

[0028] The step of providing a current difference of the pulse current;

[0029] The steps for tracking the periodic changes of current and energy storage capacitor voltage.

[0030] A computer storage medium is used to store a computer program. When the computer program is read by a computer, the computer executes the method.

[0031] The computer program product, as a computer program, implements the method described above when the computer program is read by a computer.

[0032] The present invention is beneficial in that:

[0033] 1. The effect of DC-DC converter for power decoupling

[0034] When a pulse load is connected, a DC-DC converter connected in parallel with the DC output terminal is used, and a buck-boost bidirectional converter is adopted to decouple the fluctuation component of the pulse load so that the system only needs to output the average value of the pulse current.

[0035] Beneficial effects:

[0036] Through power decoupling, the impact of current fluctuations caused by pulse loads on the DC side voltage of the system is reduced, ensuring the stability of DC output.

[0037] Comparison of existing technologies: Compared with the traditional method of directly using output capacitors to absorb current fluctuations, DC-DC converters can significantly reduce the volume and weight of output capacitors because they transfer the pulsating component to the converter for processing, reducing dependence on capacitor capacity and improving the compactness and response speed of the system.

[0038] 2. The synergistic effect of energy storage capacitors and DC-DC converters

[0039] The energy storage capacitor is connected in parallel with the DC-DC converter to provide a current difference when a pulse load is connected, ensuring that the system outputs a stable average current.

[0040] Beneficial effects:

[0041] Power regulation and stability: By absorbing and releasing excess pulse current through energy storage capacitors, the system can still output average current stably when the pulse load starts and stops, reducing the voltage fluctuation of the system and improving the power supply quality.

[0042] Comparison with existing technologies: Compared with traditional closed-loop control, the design of this energy storage capacitor makes the system more flexible in responding to changes in pulse power, thereby significantly reducing the adverse effects of voltage transient changes.

[0043] 3. Effect of combining model predictive control (MPC) with fuzzy control

[0044] Model predictive control (MPC) is used to track the reference current of the DC-DC converter in real time. At the same time, fuzzy control is combined to adaptively adjust the weight coefficient in MPC online.

[0045] Beneficial effects:

[0046] Real-time tracking and response speed: Model predictive control predicts the system state in the next switching cycle and adjusts the reference current of the DC-DC converter in real time, enabling the system to respond quickly to changes in pulse loads and ensuring the accuracy of current tracking.

[0047] Reduce computational complexity: The fuzzy controller adaptively adjusts the MPC weight coefficients and dynamically adjusts the control priority through weight adaptation, which reduces the amount of complex calculations and makes the control system more adaptable to rapidly changing load requirements.

[0048] Comparison of existing technologies: Compared with traditional proportional integral (PI) control, MPC combined with fuzzy control not only improves the tracking accuracy of the reference current, but also effectively solves the problem of slow response of PI control when the load changes drastically, thereby ensuring the periodic stability of current and capacitor voltage, and improving the decoupling effect and system robustness.

[0049] 4. Effect of pulse current decomposition strategy

[0050] The pulse current is decomposed into average current and difference current. The system only outputs the average current, and the difference current is compensated by the DC-DC converter and the energy storage capacitor.

[0051] Beneficial effects:

[0052] Enhanced system stability: Through this decomposition and compensation strategy, the impact of pulse load on the system can be reduced, the system voltage can be kept within a reasonable fluctuation range, and the stability of the DC side voltage can be ensured.

[0053] Improve power quality: This compensation strategy ensures that the current waveform distortion of the system is minimized when the pulse load is connected, improves the power factor of the three-phase AC power supply, and improves the power quality of the power supply system.

[0054] Comparison with existing technologies: Compared with the traditional method of directly responding to pulse loads, this decomposition and compensation solution effectively reduces the voltage changes in the system caused by large current fluctuations, and improves the overall stability and safety of the power supply system.

[0055] 5. The adaptive adjustment effect of fuzzy control on weight coefficient

[0056] Fuzzy control is added to MPC to adaptively adjust the weight coefficients to cope with sudden changes in pulse loads.

[0057] Beneficial effects:

[0058] Enhance system adaptability: By adaptively adjusting the weight coefficient, the controller can respond quickly to rapid load changes and maintain system stability. Especially in the case of sudden increase or decrease of pulse load, it can ensure rapid tracking and precise control of current.

[0059] Improve compensation accuracy: The dynamic adjustment of fuzzy control weights allows different error sources (current error, capacitor voltage error) to have different priorities during the compensation process, thereby achieving better compensation effects.

[0060] Comparison with existing technologies: Compared with MPC control with fixed weight coefficients, fuzzy adaptive regulation can significantly improve the response speed and compensation effect to sudden loads, avoiding the system response delay problem caused by fixed weights.

[0061] It is suitable for use in pulse load power compensation. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] Figure 1 It is a schematic diagram of the overall topology of the Y-type rectifier and the DC-DC converter;

[0063] Figure 2 This is the waveform diagram of the DC voltage in the real object when the DC side is connected to a pulse load;

[0064] Figure 3 This is the waveform diagram of the three-phase current on the AC side when the DC side is connected to a pulse load;

[0065] Figure 4 The pulse current and average current waveform diagram theoretically generated when the pulse load is connected to the DC side;

[0066] Figure 5 The waveform diagram of the theoretical tracking effect of the DC-DC converter reference current;

[0067] Figure 6 It is a schematic diagram of the DC-DC converter topology;

[0068] FIG. 7( a ) is a waveform diagram of the current error membership function; FIG. 7( b ) is a waveform diagram of the voltage error membership function; FIG. 7( c ) is a waveform diagram of the current area error membership function;

[0069] FIG8(a) is a membership function waveform diagram of weight coefficient λ1; FIG8(b) is a membership function waveform diagram of weight coefficient λ2;

[0070] Fig. 9 This is the fuzzy prediction control flow chart used in the compensation circuit;

[0071] Figure 10 shows the energy storage inductor current i under PI control and fuzzy predictive control of the DC-DC converter. Ls Waveform diagram in real object;

[0072] Fig.11 The voltage V of the energy storage capacitor under PI control and fuzzy predictive control of the DC-DC converter is Cs Waveform diagram in real object;

[0073] Fig.12 It is a schematic diagram of the waveform of the DC side voltage in the real object under the fuzzy predictive control of the DC-DC converter;

[0074] Fig.13 When working with pulse load, the waveform of three-phase current in the real object after adding compensation circuit on the DC side;

[0075] Fig.14 When working under pulse load, the harmonic analysis diagram of the three-phase current after adding the compensation circuit on the DC side. DETAILED DESCRIPTION

[0076] In order to make the advantages and benefits of the technical solution provided by the present invention more clearly reflected, the technical solution provided by the present invention is now further described in detail with reference to the accompanying drawings, specifically:

[0077] Implementation method 1: This implementation method aims to solve the above technical problems through a pulse load compensation strategy based on fuzzy predictive control, propose a more efficient and accurate control method, reduce the impact of pulse load on the DC side of the system, and achieve improved power quality and stable operation of the system.

[0078] The technical solution of this embodiment is a pulse load compensation strategy based on fuzzy predictive control, which aims to reduce the impact of pulse load on the DC side voltage of the system. The solution includes the following main parts:

[0079] System circuit: three-phase AC power supply, Y-type rectifier, output capacitor, DC-DC converter, energy storage capacitor, load, etc.

[0080] Compensation strategy: By decomposing the pulse load into average current and difference current, the DC-DC converter is used to decouple the output capacitor and load to achieve stable system output.

[0081] Model Predictive Control (MPC): Combined with fuzzy control, the weight coefficients are adaptively adjusted to improve the accuracy of current tracking.

[0082] The specific steps include:

[0083] 1. Circuit structure construction: Build a system structure including three-phase AC power supply, Y-type rectifier, output capacitor, DC-DC converter, etc.

[0084] 2. Pulse load characteristic analysis: Analyze the periodicity and low-frequency characteristics of the pulse load and calculate the average value of the pulse current.

[0085] 3. Implementation of compensation strategy: The current difference is used as the reference current of the DC-DC converter and the reference value is tracked through model predictive control.

[0086] 4. Fuzzy control optimization: Fuzzy control is added to model predictive control to optimize current tracking and capacitor voltage stability.

[0087] Detailed description of each step

[0088] Circuit structure construction

[0089] The circuit structure includes: a three-phase AC power supply, a bias voltage source, a Y-type rectifier, an output capacitor, a DC-DC converter and an energy storage capacitor.

[0090] AC power supply: provides power to the entire system.

[0091] Y-type rectifier: converts three-phase AC power into DC power, and connects the DC side to ordinary load or pulse load.

[0092] Output capacitor: connected in parallel with the DC output terminal to absorb current disturbances and stabilize the voltage of ordinary loads.

[0093] DC-DC converter: A buck-boost bidirectional converter is used to achieve power decoupling when pulse loads are connected.

[0094] Energy storage capacitor: connected in parallel with the DC-DC converter to provide the current difference required when the pulse load is connected to the DC end.

[0095] Pulse load characteristics analysis

[0096] Pulse load characteristics: Pulse load has periodic and low-frequency characteristics, with large instantaneous power changes, high peak power, but low average power.

[0097] Average current calculation: According to the periodicity of the pulse load, the average current value within one cycle of the pulse current is calculated. The system outputs the average current, and the difference is provided by the parallel compensation circuit.

[0098] Current decomposition: The pulse current is divided into average current and difference current so that the control system can stably output the average current while reducing the impact of the pulse current on the DC side voltage.

[0099] Compensation strategy implementation

[0100] Model Predictive Control (MPC):

[0101] Reference current tracking: The difference between the pulse current and its average value is used as the reference current of the DC-DC converter to control the on and off of the switch to achieve real-time tracking of the reference current.

[0102] Improved system stability: By controlling the DC-DC converter through MPC, the system only needs to output average current, thereby reducing the impact on the DC side voltage and three-phase current waveform.

[0103] Fuzzy Control Optimization

[0104] Current tracking error and capacitor voltage control:

[0105] Fuzzy control is used to adaptively tune the weight coefficients in MPC online to achieve more accurate current tracking.

[0106] Fuzzy controller design: According to the particularity of pulse load, when the current error exceeds the preset value, the corresponding weight coefficient is increased first to make the current tracking faster.

[0107] Cost function optimization: Define the cost function and take the error between the predicted current and the reference current and the error of the energy storage capacitor voltage as the optimization target to obtain the optimal switching state and achieve power decoupling.

[0108] Complete step summary and implementation details

[0109] 1. Build the circuit structure and determine how to connect the components of the system.

[0110] 2. Analyze the pulse load characteristics, calculate the average value of the pulse current, and determine the current difference that needs to be compensated.

[0111] 3. Implement compensation strategy, use DC-DC converter to track the current difference of pulse load, and perform real-time control through MPC to ensure stable current output.

[0112] 4. Optimize controller performance and use fuzzy controller to adaptively adjust the weight of MPC, improve current tracking accuracy and maintain the periodic change of energy storage capacitor voltage to ensure the decoupling effect and stability of the system.

[0113] This solution minimizes the impact of pulse load on the DC side through fuzzy predictive control to ensure stable operation of the system. The output of each step is used as the input of the next step to form a closed-loop control system, which improves the system's adaptability and response speed to pulse loads.

[0114] Implementation method 2: This implementation method successfully solves the system stability problem caused by pulse load access through the cooperation of multiple innovative technical means. Compared with other existing studies, this solution introduces key innovations such as DC-DC converter power decoupling, compensation of energy storage capacitors, MPC combined with fuzzy control, etc., which significantly improves the system's stability, power regulation capability, and response speed to pulse loads, thereby better meeting the power supply requirements under complex pulse loads in practical applications.

[0115] The following is a detailed description of the pulse load compensation strategy based on fuzzy predictive control:

[0116] It mainly includes the construction of circuit structure, analysis of pulse load characteristics, implementation of compensation strategy, and optimization of control method.

[0117] 1. Build the circuit structure

[0118] A circuit system including three-phase AC power supply, Y-type rectifier, output capacitor, DC-DC converter, energy storage capacitor and other components is constructed for pulse load power compensation.

[0119] Detailed description:

[0120] 1.1 Three-phase AC power supply: The system is powered by a three-phase AC power supply to provide electrical energy for the entire compensation system.

[0121] 1.2Y-type rectifier: The three-phase AC power supply is connected to the input end of the Y-type rectifier, which converts the AC power into DC power and transmits it to the DC side output end.

[0122] 1.3 Output capacitor: The output capacitor is connected in parallel with the DC output terminal and is mainly used to absorb current disturbances under normal load conditions to ensure the stability of the DC voltage.

[0123] 1.4 DC-DC converter: The DC-DC converter adopts a buck-boost bidirectional converter, which is connected in parallel with the DC output terminal to achieve power decoupling when the pulse load is connected to ensure that the system compensates for the pulse current fluctuation component.

[0124] 1.5 Energy storage capacitor: The energy storage capacitor is connected in parallel with the DC-DC converter and is responsible for providing the difference in system current when the pulse load is connected, ensuring the stability of the system voltage during transient changes.

[0125] The complete circuit structure can convert and compensate the input electrical energy, providing a basis for subsequent pulse load compensation and control.

[0126] 2. Pulse load characteristics analysis

[0127] Analyze the characteristics of pulse loads, clarify their impact on the system, and determine compensation strategies.

[0128] Detailed description:

[0129] 2.1 Pulse load characteristics: Pulse load has low frequency and periodicity, and is characterized by rapid changes in instantaneous power, high peak power, and low average power, and the pulse period is usually tens to hundreds of milliseconds.

[0130] 2.2 Average current calculation: According to the periodic characteristics of the pulse load, the average value of the pulse current in one cycle is calculated so that the system only needs to output the average current to reduce the impact on the system voltage.

[0131] 2.3 Current decomposition: The pulse current is decomposed into average current and difference current. The subsequent compensation circuit will compensate these differences through the control system to maintain system voltage stability.

[0132] The average value and difference of the pulse current will be used as the input of the subsequent compensation circuit to guide how to adjust and compensate the pulse current.

[0133] Implementing compensation strategies

[0134] The pulse load is decoupled through the DC-DC converter, and the energy storage capacitor provides the current difference to achieve stable output of the system.

[0135] Detailed description:

[0136] 3.1 DC-DC converter power decoupling: The DC-DC converter is connected in parallel with the DC output terminal, and a buck-boost bidirectional converter is used. Its main function is to decouple the fluctuating component of the pulse current from the system, so that the system only needs to provide the average part of the pulse current, and the fluctuating part is processed by the DC-DC converter.

[0137] 3.2 Synergistic effect of energy storage capacitor: The energy storage capacitor is connected in parallel with the DC-DC converter to provide an instantaneous current difference when the pulse load is connected. The energy storage capacitor absorbs the part of the pulse current that is higher than the average value and releases the part that is lower than the average value to ensure that the output current always remains at a stable level.

[0138] 3.3 Output current regulation: Through the switch tube control of the DC-DC converter, the difference of the pulse current is used as the reference current for current tracking to ensure that the system outputs a stable average current.

[0139] The compensation circuit outputs a stable average current, while eliminating the instantaneous current fluctuation of the pulse load and reducing the impact on the system DC voltage.

[0140] 4. Control method optimization

[0141] Model predictive control (MPC) and fuzzy control are used to optimize the control of DC-DC converters to improve the response speed and stability of the system.

[0142] Detailed description:

[0143] 4.1 Model Predictive Control (MPC): A mathematical model is established for the DC-DC converter to discretize the continuously changing current and system state and predict the system state in the next switching cycle. By comparing the predicted current value with the reference current, the optimal duty cycle of the switch is calculated to achieve the goal of power decoupling.

[0144] 4.2 Adaptive adjustment of weight coefficient by fuzzy controller: In order to enhance the adaptability of MPC in rapid load changes, the fuzzy controller is used to adaptively adjust the weight coefficient in MPC online. When the change of pulse load exceeds the preset value, the weight coefficient of current error is increased first to ensure rapid current tracking and maintain system stability.

[0145] 4.3 Definition of cost function: The difference between the predicted current and the reference current and the change in the energy storage capacitor voltage are used as parameters of the cost function. The switching state is optimized by minimizing the cost function so that the system can achieve accurate current tracking and periodic voltage change control.

[0146] The optimized control strategy enables the system to respond more quickly to pulse load changes and ensures the stability of current and voltage, ultimately improving the reliability and power quality of the system.

[0147] Implementation Method 3: Combination Figure 1-14 This embodiment is described. This embodiment is the technical solution provided above, and provides specific examples and analysis processes, specifically:

[0148] The purpose of this embodiment is to propose a pulse load compensation strategy based on fuzzy predictive control. This strategy starts from the pulse power required when the pulse load appears, and converts it into a pulse current. According to the low-frequency and periodic characteristics of the pulse load, it is deduced that the pulse current also has periodicity and low frequency, so as to calculate the average current value of the pulse current within one cycle, so that the system can stably output its current average value. The part of the pulse current exceeding the instantaneous value is provided by the parallel compensation circuit. When the instantaneous value of the pulse current is lower than the average current value, the parallel compensation circuit absorbs the excess current to stabilize the system.

[0149] A pulse load compensation strategy based on fuzzy predictive control, the circuit structure is as follows Figure 1 As shown, it includes: a three-phase AC voltage source V abc , DC bias voltage source U pz , AC side input inductor L, Y-type rectifier, DC side output capacitor C o , load, DC-DC converter, energy storage capacitor C s When the DC side is connected to a normal load and the DC-DC converter does not adopt a compensation strategy, under normal closed-loop control, the DC side output voltage can be stabilized within the allowable range of the reference voltage. The waveform of the DC side voltage in the real object is as follows: Figure 2 As shown, at 1.5s, the normal load is switched to the pulse load. It can be found that after 1.5s, the DC side voltage fluctuates greatly and cannot change within the allowable error range of the reference voltage, which does not meet the actual requirements. Figure 3 This is the waveform diagram of the three-phase current on the AC side when the pulse load is connected to the DC side in this implementation mode. It can be seen that after the pulse load is connected for 1.5s, the three-phase current waveform is significantly distorted, which does not meet the engineering requirements. Since the pulse load has a low operating frequency and a fast start-stop speed, it is difficult for traditional closed-loop control to respond to changes in the pulse load power in a timely manner; under ideal conditions, the DC side voltage fluctuates within a certain range of the reference voltage. According to the relationship between voltage and load, the load current is also a pulse current, so the pulse current also has periodicity, such as Figure 4 As shown, let the maximum value of the pulse current be I high , the minimum value is I low , the average value is I pj Since the pulse current appears and disappears almost instantly, it will have a huge impact on the entire system and affect the stability of the system. In order to reduce the impact of the pulse load on the entire circuit, the Y-type rectifier can be controlled to only provide the load pulse current i o The DC component I pj ,i o The pulsating component in the output capacitor C o Directly provide; if the switching component is ignored, let the Y-type rectifier output current be i o ′, the output capacitor current is i Co , T p is the pulse period. If the pulsating component is provided by the output capacitor and the Y-type rectifier only provides the DC current in the load current, then the corresponding output voltage pulse peak drop V o_dr for:

[0150]

[0151] Then, the output energy storage capacitor value can be expressed as:

[0152]

[0153] In the formula, f p =1 / T p , is the pulse frequency. From the above, we can see that when the output voltage pulse drops V o_dr When D p The larger the f p The lower the C o The larger the output capacitance C o As an energy storage unit that provides the pulsating component of the output power, its capacitance is very large, resulting in a large volume and weight of the capacitor. In order to reduce the volume and weight of the output capacitor, this patent places the DC-DC converter on the energy storage capacitor C. s and the load, so that the DC side of the entire system only needs to output an average current I pj When the pulse load appears, the system and the compensation circuit output current at the same time. The output current of the compensation circuit is the difference between the pulse current and the average current to meet the current requirements of the pulse load. When the pulse load disappears, the system output current I pj The excess current is absorbed by the compensation circuit to keep the system stable. The reference current i of the compensation circuit is obtained by subtracting the average current output by the system from the instantaneous value of the pulse current. Ls * , the theoretical tracking effect is as follows Figure 5 As shown; in order to achieve the compensation effect, the compensation circuit selects a DC-DC converter, which is a bidirectional buck-boost converter, such as Figure 6 As shown; the converter has two switching tubes T D1 , T D2 , using the complementary conduction strategy, when T D1 When conducting, according to Kirchhoff's voltage and current laws, the mathematical equation can be expressed as:

[0154]

[0155]

[0156] Discretizing the above expression using Euler, we can get:

[0157]

[0158] When T D2 When it is turned on, the mathematical expression can also be obtained as follows:

[0159]

[0160] After discretization, we get:

[0161]

[0162] V Cs (k+1)=V Cs (k)

[0163] The state space model of the two switching states can be derived as:

[0164] X(k+1)=A i X(k)+B i v

[0165] Y(k+1)=C i X(k)

[0166] In the formula, X(k+1)=[i Ls (k),V Cs (k)] T , Y(k) is the voltage of the energy storage capacitor, and v is the input voltage of the converter. i , B i and C i are system matrices of two switching states respectively. For the current tracking of this embodiment, an intuitive method is to take the absolute value of the difference between the given current and the current predicted by the model as the cost function. Under ideal conditions, due to the periodicity of the pulse current, the energy storage capacitor C s The upper voltage value will also change periodically with the pulse current. In practice, considering that the inductor current cannot achieve zero error tracking and the influence of environmental noise, if the voltage on the capacitor is not controlled, it will cause the capacitor voltage to be unbalanced and the actual decoupling effect is not ideal. Therefore, the absolute value of the difference between the capacitor voltage reference value and the predicted value is added to the original cost function, and its expression is:

[0167] J=λ1|i Ls (k+1)-i Lsref |+λ2|V Cs (k+1)-V Csref |

[0168] In the formula, λ1 and λ2 are weight coefficients. The optimal switching state is obtained by calculating the minimum value of the cost function, thereby obtaining the optimal solution. At the same time, in order to ensure the accuracy of current tracking, this patent uses the area equivalence principle to calculate the area S(k+1) between the reference current and the time axis in the next switching cycle based on mathematical modeling. D1 When it is on, its expression is:

[0169]

[0170] Simplified:

[0171]

[0172] T D2When conducting, it is the same as the above formula, simplified to:

[0173]

[0174] The difference between the predicted current and the area of ​​the time axis is added to the cost function of the model predictive control, and the expression is as follows:

[0175] J=λ1|i Ls (k+1)-i Lsref |+λ2|V Cs (k+1)-V Csref |+λ3|S(k+1)-S ref |

[0176] In the formula, λ3 is the weight coefficient of the absolute value of the area. By adjusting the weight coefficient, the inductor current I Ls More accurately track the reference current while ensuring the periodic change of the voltage on the energy storage capacitor; by adjusting the weight coefficient, the inductor current I Ls Track the reference current more accurately while ensuring the periodic change of the voltage on the energy storage capacitor; if only the traditional MPC controller with fixed weight coefficient is used for current tracking, when the reference current changes widely during the start and stop of the pulse, due to the limitation of the capacitor reference voltage cost function in the cost function, the tracking speed of the actual current is slow during the start and stop of the pulse cycle, affecting the decoupling effect of the system. This patent designs an MPC controller with fuzzy adaptive adjustment of the weight coefficient. According to the particularity of the pulse load, the weight coefficient priority is set to the highest after the current error exceeds the preset value. The designed controller adaptively adjusts the weight coefficient in the cost function online according to the difference between the reference current and the predicted current to ensure that the actual current can be quickly tracked when the reference current changes suddenly, so as to meet the compensation effect.

[0177] According to the principle of fuzzy control, current tracking error, capacitor voltage tracking error and area tracking error are defined. Each error is multiplied by a certain gain as input to ensure that each error can have an effect on the cost function. The corresponding weight coefficients of the above three in the cost function are defined as output. Combined with the characteristics of the pulse load, four fuzzy sets are designed for the absolute value of the current error: S (small), M (medium), MB (medium-large), and B (large). The membership function is shown in Figure 7(a). Five fuzzy sets are designed for the absolute value of the voltage error: S (small), SM (small-medium), M (medium), MB (medium-large), and B (large). The membership function is shown in Figure 7(b). Three fuzzy sets are designed for the absolute value of the area error: S (small), M (medium), and B (large). Large), and its membership function is shown in Figure 7(c); five fuzzy sets are designed for the three output weight coefficients: S (small), SM (small medium), M (medium), MB (medium large), and B (large), and its membership function is shown in Figures 8(a) and (b); the rules of fuzzy reasoning are set according to expert experience. The main purpose of this compensation circuit is to ensure real-time current tracking. For the rapid drop and rise of the reference current, the current and current area weight coefficients in the cost function are set to the maximum, giving it the highest priority, so that the current can be tracked quickly. When the current error is less than the preset value, the voltage weight coefficient in the cost function is appropriately increased through fuzzy control to achieve the purpose of voltage tracking; the fuzzy rules of weight coefficients λ1, λ2, and λ3 are shown in Tables 1 and 2;

[0178] Table 1 Fuzzy rules of λ1 / λ3

[0179]

[0180] Table 2 Fuzzy rules of λ2

[0181]

[0182] The implementation process of fuzzy predictive control in this embodiment is as follows: Fig. 9 As shown, the above three errors are used as fuzzy control inputs to obtain corresponding weight coefficients. By predicting the system state of the next switching cycle, the optimal switching duty cycle is obtained to achieve pulse load power decoupling.

[0183] FIG10(a) shows the tracking of the reference current by the DC-DC converter in this embodiment after adopting fuzzy predictive control. It can be seen that compared with the ordinary PI control in FIG10(b), the error between the tracking technology proposed in this embodiment and the reference current is smaller, and Fig.11 The voltage of the energy storage capacitor V Cs It can be seen from the waveform diagram in the actual object that the fuzzy predictive control in this embodiment can not only make the current tracking more accurate, but also ensure the periodic change of the energy storage voltage, continuously decouple the pulse load, and keep the system stable. Fig.12The waveform of the DC side voltage in the real object after the compensation circuit is added with fuzzy predictive control is shown. It can be found that after adding current compensation, the voltage can be within the allowable error range of the project, meeting the design requirements. Fig.13 The waveform diagram of the three-phase current in the real object after adding fuzzy predictive control to the compensation circuit is shown in Figure 2. Fig.14 A THD (total harmonic distortion) analysis was performed on it and it was found that its distortion rate was less than 5%, which met the engineering requirements and verified the practicality and correctness of this implementation method.

[0184] The technical solution provided by the present invention is further described in detail above through several specific implementation modes in order to highlight the advantages and benefits of the technical solution provided by the present invention. However, the several specific implementation modes described above are not intended to be used as limitations on the present invention. Any reasonable modification and improvement of the present invention, combination of implementation modes and equivalent substitution within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A pulse load compensation device based on fuzzy predictive control, characterized in that: The device comprises: Three-phase AC power supply for providing electrical energy; Y-type rectifier, used to convert three-phase AC power into DC power; Output capacitor, used to absorb current disturbances and maintain voltage stability under normal load; DC-DC converter, used to achieve power decoupling when pulse load is connected; Energy storage capacitor, used to provide the current difference when the pulse load is connected to the DC terminal; The control module is used to control the DC-DC converter, track the reference value of the pulse current in real time, realize current tracking, and reduce the impact of the pulse load on the DC side voltage of the system.

2. The pulse load compensation device based on fuzzy predictive control according to claim 1 is characterized in that: The input end of the Y-type rectifier is connected to the three-phase AC power supply, and the output end is connected to the DC output end.

3. The pulse load compensation device based on fuzzy predictive control according to claim 2 is characterized in that: The output capacitor is connected in parallel with the DC output terminal.

4. The pulse load compensation device based on fuzzy predictive control according to claim 2 is characterized in that: The DC-DC converter is connected in parallel with the DC output terminal.

5. The pulse load compensation device based on fuzzy predictive control according to claim 4 is characterized in that: The DC-DC converter is implemented by a buck-boost bidirectional converter.

6. The pulse load compensation device based on fuzzy predictive control according to claim 1 is characterized in that: The energy storage capacitor is connected in parallel with the DC-DC converter.

7. The pulse load compensation device based on fuzzy predictive control according to claim 1 is characterized in that: The control module tracks the reference value of the pulse current in real time through a model prediction controller and a fuzzy controller, and realizes the tracking of the periodic changes of the current and the energy storage capacitor voltage through adaptive online adjustment of the weight coefficient.

8. A pulse load compensation method based on fuzzy predictive control, characterized in that: The method is implemented based on the device according to claim 1, and includes: Steps to convert AC to DC; The steps to absorb the current disturbance of DC power and maintain the voltage stability under normal load; The step of decoupling the power when the pulse load is connected; The step of providing a current difference of the pulse current; The steps for tracking the periodic changes of current and energy storage capacitor voltage.

9. A computer storage medium for storing a computer program, characterized in that: When the computer program is read by a computer, the computer executes the method of claim 8.

10. A computer program product, being a computer program, characterized in that When the computer program is read by a computer, the method according to claim 8 is implemented.

Images