LC filter-based cascaded fuzzy feedforward direct-drive speed regulation control method and system

Abstract

This invention relates to a cascaded fuzzy feedforward direct-drive speed control method and system based on LC filtering. The method includes: obtaining a speed error and error rate of change from a reference speed and a real-time measured speed; inputting the speed error and error rate of change into a Mamdani fuzzy inference engine to output a power setpoint coefficient; inputting the power setpoint coefficient and a normalized measured speed signal into a two-dimensional torque-current reference lookup table, performing bilinear interpolation calculation, and outputting an electromagnetic torque command; converting the electromagnetic torque command into a quadrature-axis command current, generating a PWM signal through current closed-loop regulation and SVPWM modulation, and driving a two-level voltage-type inverter; the output of the two-level voltage-type inverter drives a low-voltage permanent magnet synchronous motor through an LC sine wave filter with an RC damping branch. Compared with the prior art, this invention has advantages such as low computational load, improved dynamic response, and enhanced disturbance rejection.

Description

Technical Field

[0001] This invention relates to the field of electrical drive and intelligent motor control technology, and in particular to a cascaded fuzzy feedforward direct drive speed control method and system based on LC filtering. Background Technology

[0002] In industrial drive systems such as mine hoists, conveyor belts, and heavy-duty pumping stations, the power supply is typically a medium-to-high voltage AC grid of 1140V, while the end-effectors often use low-voltage, high-power AC motors (such as permanent magnet synchronous motors) with a rated voltage of 380V. For this "high-voltage power supply, low-voltage drive" application scenario, existing variable frequency speed control systems suffer from the following significant physical bottlenecks and technical deficiencies in their hardware architecture and control algorithms: First, traditional power frequency step-down solutions result in large system size and low power density.

[0003] The most common practice is to install bulky power frequency or intermediate frequency step-down transformers at the front or back of the frequency converter. However, in extremely restricted underground explosion-proof spaces such as mines, transformers not only occupy a large volume, greatly reducing the overall power density and mobility of the system, but also increase additional copper and iron losses, reducing transmission efficiency.

[0004] Second, there are insulation breakdown and high-frequency surge problems caused by high-voltage direct drive.

[0005] To eliminate the need for a step-down transformer, some technologies attempt to use a topology of "high-voltage inverter directly driving low-voltage motor". After the 1140V AC power is rectified to a DC bus voltage of approximately 1600V or higher, the pulse width modulation (PWM) chopped voltage generated by the inverter's high-frequency switching is directly applied to the 380V motor windings. This extremely high voltage difference leads to a very high voltage sag rate (dv / dt) and transient surge glitches, which not only generate severe electromagnetic interference (EMI) but also rapidly damage the stator insulation system of the low-voltage motor, causing catastrophic breakdown and short circuit.

[0006] Third, the introduction of LC sine wave filters induces higher-order resonance and controlled divergence.

[0007] To smooth high-voltage PWM chopping and protect motor insulation at the physical level, inserting a sinusoidal filter (such as an LC sinusoidal filter) between the inverter and the motor is a necessary hardware measure. However, the introduction of the LC sinusoidal filter transforms the original first-order motor system into a complex high-order underdamped system. Traditional closed-loop control systems are prone to exciting the inherent resonant frequency of the LC loop, leading to severe distortion of the current waveform and even causing numerical explosion in the control algorithm and system instability.

[0008] Fourth, traditional linear PI control has extremely poor dynamic performance under heavy load and strong disturbance conditions.

[0009] At the control algorithm level, existing field-oriented control (FOC) typically uses traditional linear proportional-integral (PI) controllers as the main components of the speed and current loops. When faced with the extremely high static friction and equipment rotational inertia in industrial environments (i.e., "dead weight" start-up), linear PI controllers struggle to balance "ultra-fast response" and "no overshoot." When encountering sudden nonlinear disturbances such as large pieces of material jamming, the integral lag effect of the PI controller causes a significant drop in speed with extremely slow recovery, completely failing to meet the stringent requirements for strong burst force and rapid correction capabilities under harsh operating conditions.

[0010] In summary, existing variable frequency speed control systems urgently need a novel hardware-software co-engineering architecture that can perfectly solve the insulation threats and LC resonance problems caused by high-voltage direct drive without eliminating the need for a step-down transformer. Furthermore, it should overcome the limitations of linear PI control algorithms to achieve extremely fast dynamic response and stable control under conditions of large inertia and strong disturbances. A search revealed Chinese invention patent application publication number CN110971169B, which discloses a direct torque control method for permanent magnet synchronous motors based on fuzzy output duty cycle. First, a voltage vector is selected using a switching meter. Then, three input quantities—the current torque error, the rate of change of torque error, and the stator flux linkage error—are input into a fuzzy controller. These input quantities undergo fuzzification, fuzzy inference, and defuzzification within the fuzzy controller, outputting the duty cycle corresponding to the selected voltage vector. This output is then used for direct torque control via space vector modulation. The fuzzy controller is only used for online correction of control parameters; the torque command is directly output from the speed PI loop. Constraint equations are solved online in segments for the constant torque region and the field weakening region. This existing patent application suffers from problems such as high computational load, slow dynamic response, and susceptibility to fluctuations during operating condition switching.

[0011] The technical problem that needs to be solved is how to eliminate the need for a step-down transformer, enable high-voltage power supply to directly drive low-voltage motors, and solve the resulting problems of insulation threats, LC resonance, and poor dynamic response under strong disturbances. Summary of the Invention

[0012] The purpose of this invention is to overcome the defects of the prior art and provide a cascaded fuzzy feedforward direct drive speed control method and system based on LC filtering.

[0013] The objective of this invention can be achieved through the following technical solutions: According to one aspect of the present invention, a cascaded fuzzy feedforward direct drive speed control method based on LC filtering is provided, the method comprising: The measured speed signal of the low-voltage permanent magnet synchronous motor is acquired in real time, and the speed error and error change rate are obtained from the reference speed and the measured speed. The speed error and the rate of change of error are input into the Mamdani fuzzy inference engine, and the output force given coefficient is output. The output given coefficient and the normalized measured speed signal are input into a two-dimensional torque-current reference lookup table, bilinear interpolation calculation is performed, and an electromagnetic torque command is output; the two-dimensional torque-current reference lookup table has a built-in bilinear interpolation algorithm and discrete electromagnetic torque grid data; The electromagnetic torque command is converted into quadrature axis command current, which is then used to generate a PWM signal through current closed-loop regulation and SVPWM modulation to drive a two-level voltage-type inverter. The output of the two-level voltage source inverter drives a low-voltage permanent magnet synchronous motor via an LC sine wave filter with an RC damping branch.

[0014] As a preferred technical solution, the process of obtaining the output given coefficient includes: The speed error and the rate of change of error are multiplied by the quantization gain and mapped to the fundamental universe of discourse. Fuzzy inference is performed based on preset fuzzy sets and inference rules; The fuzzy inference results are defuzzified and amplitude-limited to obtain the output given coefficient.

[0015] As a preferred technical solution, the fuzzy inference based on preset fuzzy sets and inference rules specifically includes: Within the basic domain, the rotational speed error and the rate of change of error are both divided based on a preset fuzzy set; the preset fuzzy set is {NB, NM, NS, ZE, PS, PM, PB}. An asymmetric rule matrix is ​​formed by the intersection of fuzzy sets of speed error and error change rate. Its boundary mapping conditions include: when the input speed error is NB and the error change rate is NB, the fuzzy set of the mapping output is NS; when the input speed error is PB and the error change rate is PB, the fuzzy set of the mapping output is PB.

[0016] As a preferred technical solution, the discrete electromagnetic torque grid data simultaneously satisfies the current-limiting circle constraint and the voltage-limiting ellipse constraint. The discrete electromagnetic torque grid data corresponds to the analytical solution of the maximum torque-to-current ratio control criterion in the constant torque region below the base speed, and to the analytical solution of the torque maximization control criterion under voltage constraints in the weak magnetic region above the base speed.

[0017] As a preferred technical solution, the output given coefficient and the normalized measured speed signal are used as coordinate input values. Bilinear interpolation calculation is performed based on the bilinear interpolation algorithm in the two-dimensional torque current reference lookup table to output the electromagnetic torque command. The electromagnetic torque command is converted into the cross-axis command current through torque-current gain conversion. Simultaneously, the normalized measured speed signal is input into a one-dimensional torque current reference lookup table, and the direct-axis command current for field weakening control is output after the lookup table is consulted.

[0018] As a preferred technical solution, the generation of the PWM signal through current closed-loop regulation and SVPWM modulation includes: The feedback current of the low-voltage permanent magnet synchronous motor is obtained. The difference between the quadrature axis command current, the direct axis command current and the corresponding feedback current is calculated. The two difference calculation results are input into the corresponding discrete-time PID controllers, and after independent adjustment, AC voltage command and DC voltage command are output respectively. The AC voltage command and DC voltage command are amplitude-limited and then input into the SVPWM generator, where they are modulated to output a PWM signal.

[0019] As a preferred technical solution, the LC sine wave filter with RC damping branch includes a three-phase filter inductor, a three-phase filter capacitor, and an RC damping branch. The three-phase filter inductor is connected in series between the two-level voltage type inverter and the low-voltage permanent magnet synchronous motor; The three-phase filter capacitors are connected in parallel between the phases on the output side of the three-phase filter inductor; The RC damping branch is connected in parallel with the three-phase filter capacitor, and within the RC damping branch, the damping resistor and the damping capacitor are connected in series.

[0020] According to another aspect of the present invention, a cascaded fuzzy feedforward direct drive speed control system based on LC filtering is provided, the system comprising a main power circuit link and a cascaded control system; The main power circuit link includes a two-level voltage source inverter and an LC sine wave filter with an RC damping branch. The cascaded control system includes a subtractor, a cascaded fuzzy decision module, a dynamic mapping execution module, a current command calculation module, and a current closed-loop and limiting drive module connected in sequence. The reference speed and the real-time measured speed signal distribution of the low-voltage permanent magnet synchronous motor are input to the two input terminals of the subtractor, and the output speed error and error change rate are output. The speed error and the rate of change of error are input into the cascaded fuzzy decision module for Mamdani fuzzy inference, and the output force given coefficient is output. The output given coefficient and the normalized measured speed signal are input into the power mapping execution module, which performs bilinear interpolation calculation based on a two-dimensional torque-current reference lookup table and outputs an electromagnetic torque command; the two-dimensional torque-current reference lookup table has a built-in bilinear interpolation algorithm and discrete electromagnetic torque grid data; The current command calculation module converts the received electromagnetic torque command into quadrature axis command current. The quadrature axis command current is then adjusted by the current closed loop and the limiting drive module and SVPWM modulation to generate a PWM signal to drive the two-level voltage type inverter. The output of the two-level voltage source inverter drives a low-voltage permanent magnet synchronous motor via an LC sine wave filter with an RC damping branch.

[0021] As a preferred technical solution, the power mapping execution module includes a two-dimensional torque-current reference lookup table and a one-dimensional torque-current reference lookup table. The two input terminals of the two-dimensional torque current reference lookup table are respectively connected to the output terminal of the cascaded fuzzy decision module and the normalization terminal of the measured speed signal, and the output terminal is connected to the input terminal of the current command calculation module. The output of the current command calculation module is connected to the current closed-loop and limiting drive module. The input of the one-dimensional torque current reference lookup table is connected to the normalized result of the measured speed signal, and the output is connected to the current closed-loop and limiting drive module.

[0022] As a preferred technical solution, the two-dimensional torque-current reference lookup table incorporates a bilinear interpolation algorithm and discrete electromagnetic torque grid data, wherein the discrete electromagnetic torque grid data simultaneously satisfies the current-limiting circle constraint and the voltage-limiting ellipse constraint.

[0023] Compared with the prior art, the present invention has the following beneficial effects: 1) This invention uses the Mamdani fuzzy inference engine to directly infer the output coefficient based on the speed error and the rate of change of the error, replacing the traditional speed PI closed-loop control, effectively improving the system's dynamic response speed and anti-load disturbance capability; at the same time, it adopts a built-in bilinear interpolation algorithm and a two-dimensional torque-current reference lookup table of discrete electromagnetic torque grid data to complete the optimal torque calculation offline, avoiding the huge amount of computation caused by the online real-time solution of constraint equations in traditional control; an LC sine wave filter with an RC damping branch is added between the inverter and the motor, which can effectively suppress the high voltage jump rate dv / dt impact of the high voltage PWM output of the two-level inverter, avoid the breakdown of the insulation of the low voltage permanent magnet synchronous motor, greatly improve the system's operational reliability and safety, and achieve the effect of unimpeded transmission of fundamental power and complete suppression of high frequency switching noise.

[0024] 2) This invention enables stable reasoning of speed error and error change rate within a unified domain through quantization mapping, fuzzy inference, and defuzzification limiting processing. This eliminates control deviations caused by dimensional inconsistencies, ensures accurate and stable output of the power setpoint coefficient, and improves the control accuracy and reliability of the fuzzy feedforward link. In the fuzzy inference, an asymmetric regular matrix composed of speed error and error change rate is used, and boundary mapping conditions are set for extreme working conditions. This can effectively suppress oscillations and overshoots under large deviations, accelerate dynamic response, improve steady-state control accuracy, and significantly improve the control performance of the motor under low-speed start-up, high-speed speed regulation, and load disturbances.

[0025] 3) The discrete electromagnetic torque grid data in the two-dimensional torque current reference lookup table of this invention simultaneously satisfies the current limit circle constraint and the voltage limit ellipse constraint; the physical constraints of current and voltage under all motor operating conditions are pre-embedded offline in the grid data, so there is no need to iterate and solve the constraint equations in real time during online operation. While ensuring that the torque setpoint strictly meets the motor's safe operation boundary and avoiding overcurrent and overvoltage faults, it significantly reduces the online computing load of the controller, realizes the optimal torque output under all operating conditions with high efficiency in the constant torque region and stable acceleration in the field weakening region, and solves the defects of traditional control with large online constraint solving calculation and large torque fluctuation during operating condition switching.

[0026] 4) The LC sinusoidal filter with RC damping branch of this invention suppresses high-frequency switching harmonics and high-voltage PWM spikes through series filter inductor and parallel filter capacitor. At the same time, the RC damping branch connected in parallel with the filter capacitor and series with the resistor and capacitor can accurately match the characteristic impedance of the LC system and suppress the inherent high-order resonance spikes of the LC, so as to achieve a smooth soft landing from the inverter's high-voltage PWM output to the motor sinusoidal voltage. Without hindering the fundamental power transmission and without generating additional energy loss, it effectively reduces the voltage jump rate, completely avoids the high-voltage spike from breaking down the insulation of the low-voltage permanent magnet synchronous motor, and significantly improves the operational stability, safety and reliability of the drive system. Attached Figure Description

[0027] Figure 1 This is a flowchart illustrating the cascaded fuzzy feedforward direct drive speed control algorithm in an embodiment of the present invention. Figure 2 This is a logic block diagram of the cascaded fuzzy feedforward direct drive speed control algorithm in an embodiment of the present invention; Figure 3 This is a schematic diagram of the structure of the cascaded fuzzy feedforward direct drive speed control system provided in an embodiment of the present invention; Figure 4 This is a simulation waveform diagram comparing the inverter output PWM voltage and the motor input voltage in an embodiment of the present invention. Figure 5 The above is a simulation waveform diagram of the dynamic response of motor speed and electromagnetic torque in an embodiment of the present invention. Detailed Implementation

[0028] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0029] To address the shortcomings of existing traditional variable frequency speed control systems, such as the need for large step-down transformers, the risk of insulation breakdown due to high-voltage direct drive, the tendency for high-order resonance divergence caused by the introduction of LC sine wave filters, and the extremely poor dynamic response of traditional linear PI control under large inertia startup and nonlinear disturbances, this invention aims to provide a cascaded fuzzy feedforward direct drive speed control method and system based on LC filtering. This invention employs a hardware-software co-design architecture to filter out high-voltage pulses at the physical hardware level to protect motor insulation, and at the software algorithm level, eliminates system resonance and improves dynamic disturbance rejection capability through cascaded fuzzy logic and mapping with physical boundary constraints.

[0030] This embodiment relates to a cascaded fuzzy feedforward direct drive speed control method based on LC filtering, such as... Figure 1 and Figure 2 ,include: S1. Acquire the reference speed signal and the measured speed signal. Perform a difference operation on the reference speed signal and the measured speed signal to obtain the speed error signal. Perform a differential operation on the speed error signal to obtain the corresponding error change rate.

[0031] S2: Input the speed error and error change rate into the Mamdani fuzzy inference engine, and output the output force setpoint coefficient; input the output force setpoint coefficient and the normalized measured speed signal into the two-dimensional torque current reference lookup table, and output the electromagnetic torque command. After gain conversion processing, the electromagnetic torque command is output as the quadrature axis command current; S3: Obtain the feedback current of the motor, calculate the difference between the quadrature shaft command current and the feedback current, input the calculation result into the discrete-time PID controller, and output the voltage command after adjustment; S4 performs voltage command limiting processing. The limited voltage command is input to the SVPWM generator and outputs a PWM signal. The PWM signal is connected to the switching control terminal of the two-level voltage source inverter. The output terminal of the two-level voltage source inverter is connected to the low-voltage permanent magnet synchronous motor via an LC sine wave filter with RC damping branch.

[0032] Furthermore, in S2, such as Figure 2 The process of inputting the rotational speed error and the rate of change of error into the Mamdani fuzzy inference engine includes: S21, multiply the rotational speed error and the rate of change of error by the quantization gain respectively, and map them to the fundamental universe of discourse [-1, 1]. S22, within the basic domain of [-1, 1], the rotational speed error, error rate of change, and output given coefficient are all divided into 7 fuzzy sets {NB, NM, NS, ZE, PS, PM, PB}, as shown in Table 1.

[0033] Table 1 The membership functions of the seven fuzzy sets are all Gaussian membership functions. The variance The constant value is 0.15, and the center point is... The set of values ​​for is {-1, -0.67, -0.33, 0, 0.33, 0.67, 1}; The Mamdani fuzzy inference engine outputs defuzzification calculations using the centroid method, and its algebraic analytical expression is: ,in To defuzzify the numerical values, This is the membership function after aggregation.

[0034] Furthermore, the internal logic of the Mamdani fuzzy inference engine is constructed from the intersection of fuzzy sets of rotational speed error and error change rate, forming a 7×7 asymmetric rule matrix; the boundary mapping conditions of the 7×7 asymmetric rule matrix include: When the input rotational speed error is NB and the error change rate is NB, the mapped output fuzzy set is NS; When the input rotational speed error is PB and the error change rate is PB, the mapped output fuzzy set is PB.

[0035] S23, Defuzzification calculation output value After being saturated and truncated to the [0, 1] interval, it is used as the output given coefficient input to the two-dimensional torque current reference lookup table.

[0036] Furthermore, in S2, the process of inputting the output given coefficient and the normalized measured speed signal into the two-dimensional torque current reference lookup table includes the following analytical steps: S24 uses the output given coefficient and the normalized measured speed signal as coordinate input values ​​to perform bilinear interpolation calculation in a two-dimensional torque current reference lookup table; the two-dimensional torque current reference lookup table contains the bilinear interpolation algorithm logic and discrete electromagnetic torque grid data.

[0037] The discrete electromagnetic torque grid data pre-stored in the two-dimensional torque-current reference lookup table satisfies the following dual physical boundary constraint equations: Current limiting circle constraint equation: ; Voltage limit elliptic constraint equation: ; in, For direct-axis current, For quadrature axis current, For the maximum allowable output current, It is a direct-axis inductor. It is a quadrature axis inductor. It is a permanent magnet flux linkage. The maximum available phase voltage peak value. Electric angular velocity; The discrete electromagnetic torque grid data corresponds to the mathematical analytical solution of the maximum torque-to-current ratio criterion in the constant torque region below the base velocity, and to the torque maximization analytical solution under the voltage limit elliptic constraint in the weak magnetic region above the base velocity.

[0038] Furthermore, in S4, the process of limiting the voltage command includes: obtaining the voltage value of the high-voltage DC bus, the amplitude cutoff boundary of the limiting process being the rated insulation withstand voltage of the low-voltage permanent magnet synchronous motor; and the voltage vector amplitude corresponding to the voltage command after the limiting process being less than or equal to the amplitude cutoff boundary.

[0039] This embodiment also relates to a cascaded fuzzy feedforward direct drive speed control system based on LC filtering, such as... Figure 3 This includes the main power circuit links and cascaded control systems.

[0040] The main power circuit link includes a three-phase uncontrolled rectifier bridge ( Figure 3 The rectifier module), DC support capacitor, and two-level voltage inverter ( Figure 3 The inverter module, LC sine wave filter with RC damping branch, and low-voltage permanent magnet synchronous motor; The AC input terminal of the three-phase uncontrolled rectifier bridge is connected to the medium- and high-voltage AC power grid, and the output terminal is connected to both ends of the DC support capacitor; the voltage across the DC support capacitor is the high-voltage DC bus voltage. The DC input terminal of the two-level voltage source inverter is connected to the two ends of the DC support capacitor, and the AC output terminal of the two-level voltage source inverter is connected to the input terminal of the LC sine wave filter with RC damping branch; the output terminal of the LC sine wave filter with RC damping branch is connected to the stator winding of the low-voltage permanent magnet synchronous motor. The cascaded control system includes a subtractor, a cascaded fuzzy decision module, a power mapping execution module, a current command calculation module, and a current closed-loop and limiting drive module connected in sequence. The subtractor receives a reference speed signal and a measured speed signal at its two inputs, respectively. The output of the subtractor is connected to the input of the cascaded fuzzy decision module. The output of the cascaded fuzzy decision module is connected to the first input of the power mapping execution module, and the normalized input of the measured speed signal is connected to the second input of the power mapping execution module. The output of the power mapping execution module is connected to the input of the current closed-loop and limiting drive module, and the output of the current closed-loop and limiting drive module is connected to the control terminal of the two-level voltage source inverter.

[0041] Furthermore, the LC sine wave filter with RC damping branch includes a three-phase filter inductor, a three-phase filter capacitor, and an RC damping branch; the three-phase filter inductor is connected in series between the two-level voltage-source inverter and the low-voltage permanent magnet synchronous motor; the three-phase filter capacitor is connected in parallel between the phases on the output side of the three-phase filter inductor. Referring to Figure 1, the RC damping branch specifically refers to the resistive element connected in series with the three-phase filter capacitor; the RC damping branch, by connecting a damping resistor in series in the circuit of each phase filter capacitor, together with the three-phase filter capacitor, forms a damped filter network to absorb the high-order resonant energy caused by high-frequency switching operations.

[0042] Furthermore, the inductance value of the three-phase filter inductor The capacitance value of the three-phase filter capacitor The resistance value of the damping resistor and the capacitance value of the damping capacitor The following parameter constraints must be satisfied: , in, This is the voltage of the high-voltage DC bus. This refers to the switching frequency of a two-level voltage source inverter. Maximum current ripple limit; capacitance value ,in The cutoff frequency of the LC sine wave filter with RC damping branch; resistance value. The resistance value of the damping resistor is equal to the characteristic impedance of the LC sine wave filter with an RC damping branch; the capacitance value... Meanwhile, the cutoff frequency Satisfying frequency constraints: ,in This is the fundamental frequency of the low-voltage permanent magnet synchronous motor.

[0043] Furthermore, the cascaded fuzzy decision module includes an input quantization unit, a fuzzification unit, a rule inference unit, a defuzzification unit, and an output saturation limiter; wherein: The input quantization unit is used to map the input error signal to the standard universe of discourse; The fuzzification unit is used to calculate the membership degree of each fuzzy set corresponding to the input quantity; The rule-based reasoning unit is used to perform logical reasoning according to preset rules and output fuzzy control quantities; The defuzzification unit is used to convert fuzzy control quantities into precise output given coefficient values. Its internal logic is a centroid method algebra operator. The output saturation limiter is used to constrain the numerical range of the final output, and its numerical cutoff interval is [0, 1].

[0044] Furthermore, the power mapping execution module includes a two-dimensional torque-current reference lookup table and a one-dimensional torque-current reference lookup table; the two input terminals of the two-dimensional torque-current reference lookup table are respectively connected to the output terminal of the cascaded fuzzy decision module and the normalization terminal of the measured speed signal, the output terminal of the two-dimensional torque-current reference lookup table is connected to the input terminal of the current command calculation module, and the output terminal of the current command calculation module is connected to the current closed-loop and limiting drive module.

[0045] Among them, the two-dimensional torque current reference lookup table uses the output given coefficient. With the normalized measured speed signal Using coordinate input, and through the execution of a bilinear interpolation algorithm, the output is an electromagnetic torque command constrained by both the voltage-limiting ellipse and the current-limiting circle. ; One-dimensional torque-current reference lookup table for measuring speed signals The input is the direct-axis command current (i.e., the direct-axis reference current component) used for field weakening control. The input of the one-dimensional torque current reference lookup table is connected to the normalization terminal of the measured speed signal, and the output of the one-dimensional torque current reference lookup table is connected to the current closed-loop and limiting drive module.

[0046] The current command calculation module is used to calculate the electromagnetic torque command based on the motor torque coefficient. Converted to the corresponding quadrature axis command current (i.e., the quadrature axis reference current component).

[0047] Furthermore, the current closed-loop and limiting drive module includes a discrete-time PID controller, an inverse Park converter, a voltage limiter, and an SVPWM generator. The input of the discrete-time PID controller receives the output signals (quadrature-axis command current and direct-axis command current) from the power mapping execution module and the feedback current signal from the motor. The output of the discrete-time PID controller is connected to the input of the inverse Park converter; the output of the inverse Park converter is connected to the input of the voltage limiter, and the output of the voltage limiter is connected to the input of the SVPWM generator. The output of the SVPWM generator outputs a PWM signal, which is connected to a two-level voltage-source inverter.

[0048] The difference between the quadrature-axis command current and the feedback current is input to the discrete-time PID controller, which outputs a voltage command. The inverse Park converter combines the voltage command with the motor rotor position angle θ to transform it back into a voltage command in a two-phase stationary coordinate system (αβ axis), which is then used by the subsequent SVPWM generator. The voltage limiter limits the voltage command, and the limited voltage command is input to the SVPWM generator, which outputs a PWM signal.

[0049] It achieves deep collaboration between software and hardware, and has the following significant benefits: 1. Completely eliminate the step-down transformer, achieving ultimate power density and efficiency: This invention adopts a topology architecture of direct rectification of medium- and high-voltage AC to drive a low-voltage motor, completely eliminating the bulky step-down transformer in traditional frequency conversion solutions, which suffers from significant copper and iron losses. This not only greatly reduces the physical size and weight of the drive system and improves the energy conversion efficiency, but also perfectly meets the stringent requirements of high power density and highly mobile deployment of the transmission system in confined industrial spaces (such as mining equipment).

[0050] 2. Constructing a hardware-software co-operational physical protection matrix to completely eliminate the hidden danger of insulation breakdown in low-voltage motors: Addressing the extremely high voltage jump rate (dv / dt) and surge impact caused by high-frequency switching operations of high-voltage DC buses (e.g., up to 1612V), this invention employs a dual mechanism of hardware filtering and software limiting. On the hardware side, an LC sine wave filter strictly adheres to frequency band constraints, physically smoothing the violent PWM chopping voltage into a low-voltage sine wave suitable for the motor. On the software side, high-voltage DC bus voltage feedback is introduced into the limiting process, and combined with the voltage limit elliptic constraint equation within the two-dimensional dynamic mapping table, the final output voltage vector is strictly locked within the motor's safe tolerance boundary. This hardware-software combined mechanism eliminates the risk of insulation breakdown from a physical level.

[0051] 3. Overcoming the bottleneck of high-order resonance and ensuring high-frequency stable control of complex LC underdamped systems: Addressing the industry-wide challenge that traditional control methods, which often introduce LC sinusoidal filters, easily induce system resonance divergence, this invention proposes a perfect decoupling damping scheme. At the hardware level, an RC damping branch with a impedance strictly equal to the LC characteristic impedance provides physical critical damping. At the software level, a 7×7 asymmetric fuzzy rule matrix automatically switches to high-smoothness damping in the steady-state range, coupled with robust tuning using a low-proportion, medium-integral current loop at the bottom layer. This scheme completely eliminates transient high-frequency oscillations caused by high voltage differentials, ensuring extremely high speed accuracy under steady-state operation.

[0052] 4. Innovative cascaded feedforward decision architecture, endowing the system with superior high-inertia start-up burst power and ultra-fast disturbance rejection performance: This invention abandons the integral lag defect easily generated by traditional linear PI speed loops, and creatively constructs a decoupled decision architecture with Mamdani fuzzy inference and two-dimensional mapping of dual physical boundaries. When facing high-inertia start-up conditions such as dead weight, the system can instantly output a large output coefficient, and directly map it to the optimal current command that satisfies the maximum torque-to-current ratio (MTPA) through bilinear interpolation, realizing a rapid increase in large torque with zero overshoot; when facing nonlinear heavy-load disturbances such as sudden mechanical jamming, the algorithm, with its extremely high error change rate sensing capability, instantly bursts electromagnetic torque within milliseconds (e.g., within 200ms), suppressing the speed drop to an extremely low level, exhibiting ultra-fast dynamic correction and disturbance rejection capabilities far exceeding those of traditional controllers.

[0053] This embodiment also relates to the direct drive application of a cascaded fuzzy feedforward speed regulation system based on LC filtering in heavy-duty transportation equipment in mines (such as large belt conveyors or water pumps with the risk of sudden jamming). The cascaded fuzzy feedforward direct drive speed regulation system includes a main power circuit link and a cascaded control system.

[0054] The mine's power grid is known to be 1140V medium-high voltage AC, while the main drive equipment is a low-voltage permanent magnet synchronous motor (PMSM) with a rated voltage of 380V and a rated power of 20 kW. The specific engineering calibration parameters of this motor are: number of pole pairs... Stator winding phase resistance Direct-axis and quadrature-axis inductors Permanent magnet magnetic flux Considering the "dead weight" characteristic of mining equipment, its mechanical rotational inertia is set to a large inertia. Viscous damping coefficient This embodiment is based on discrete time steps. The control cycle operates.

[0055] I. Construction of Hardware Main Topology and Physical Boundaries like Figure 3 As shown, the cascaded fuzzy feedforward direct drive speed control system based on LC filtering of the present invention completely eliminates the bulky step-down transformer in the main power circuit link, achieving high power density.

[0056] Establishment of the high-voltage DC bus: The 1140V three-phase AC power grid is connected to the three-phase uncontrolled rectifier bridge. After being regulated by the DC support capacitor, a high-voltage DC bus of up to about 1612V is directly established.

[0057] "Soft landing" design of LC sine wave filter: Two-level voltage source inverter (switching frequency) f sw=10 kHz) directly chops and outputs 1612V DC power. Without processing, this high-voltage PWM wave with an extremely high voltage slew rate (dv / dt) would instantly break down the insulation of a 380V motor. Therefore, in this embodiment, an LC sine wave filter with an RC damping branch is inserted in series between the inverter and the motor.

[0058] Inductor and capacitor design: To keep high-frequency switching noise out without hindering the fundamental frequency dynamics, the inductance value L must meet the following requirements. (in (This refers to the allowable current ripple limit). Based on this constraint, this embodiment selects... Matching capacitors .

[0059] Frequency constraint: the cutoff frequency of this filter From the formula The frequency was determined to be approximately 1591 Hz. This frequency perfectly meets the requirements. Constraints ( (The fundamental frequency of the motor operation) ensures that the filter neither interferes with the control system nor fails to completely smooth out high-frequency glitches.

[0060] Critical matching of the RC damping branch: This is the main part of preventing high-order resonances in the LC filter. Damping resistor It is strictly set to be equal to the characteristic impedance of the LC system, that is (Matching damping capacitor) In layman's terms, this is like adding a perfectly positioned shock absorber to a spring (LC) that is prone to bouncing, which can absorb transient voltage surges very quickly without generating unnecessary energy loss.

[0061] Hardware filtering effect confirms: such as Figure 4 The simulation waveform comparison shown shows that the inverter output (UniversalBridge outlet) is filled with intense high-frequency chopping at the 1600V level; while after passing through the above-mentioned LC sine wave filter, the voltage fed into the motor stator winding is perfectly "softened" into a smooth low-voltage sine wave, which completely protects the insulation of the 380V motor from the physical level.

[0062] II. Intelligent Decoupling of Cascaded Fuzzy Decision Layer To overcome the aforementioned slow start-up and nonlinear disturbances caused by large inertia (J=0.5), this system abandons the traditional PI velocity loop, such as... Figure 2 As shown, the Mamdani cascaded fuzzy inference engine was used: State perception and mapping: The subtractor calculates the rotational speed error and its rate of change, multiplies them by the quantization gain, and then maps them to the fundamental domain of [-1,1]. In simple terms, it scales the speed deviation of hundreds or thousands of revolutions in the real world to the "brain's thinking range" of -1 to 1.

[0063] High-fidelity fuzzification: Within the interval [-1,1], the input and output coefficients α are finely divided into 7 fuzzy sets {NB, NM, NS, ZE, PS, PM, PB} (representing from negative to positive). To ensure smooth switching and eliminate current spikes caused by command jumps, all membership functions use Gaussian functions. The variance σ is locked at 0.15, and the center point... c The values ​​are evenly distributed in {-1, -0.67, -0.33, 0, 0.33, 0.67, 1}. This mathematical arrangement ensures that adjacent states overlap precisely at a membership degree of 0.5.

[0064] The cascaded fuzzy inference engine internally contains a 7×7 asymmetric rule table, with the core boundary mapping being: 1) Overvoltage protection during high-speed braking: When both the speed error and the rate of change of error are NB (motor overspeeding and still accelerating), conventional symmetrical logic will output NB to brake sharply. However, this embodiment forces mapping to NS (gentle braking). This is because under high-voltage 1612V direct drive, sharp braking will cause the motor to pump a huge reverse current, which may damage the DC capacitor, so "intermittent braking" is necessary.

[0065] 2) High Inertia Limit Burst: When both the speed error and the rate of change of error are PB (significant motor lag), the output is PB, which gives the system the ability to withstand high inertia. Maximum starting pulling force under dead weight.

[0066] Defuzzing and amplitude limiting: A centroid method integral formula that comprehensively considers the contributions of all rules is adopted. Calculate the specific value, then forcibly truncate it to the [0,1] interval using a saturation limiter, and output the output force setpoint coefficient that only allows unidirectional traction. α .

[0067] III. Two-dimensional dynamic mapping table and physical boundary constraints The fuzzy brain calculates This is merely a "percentage of intent" and must be translated into a "real command" that conforms to the physical limits of the motor. This step is accomplished using a two-dimensional torque-current reference lookup table (2D-LUT): 1. Double Limit Defense: The mesh data embedded within the 2D-LUT is not randomly filled in, but rather obtained by solving the intersection of two physical constraints offline. a) Current limiting circle: (To ensure that the inverter's IGBT transistors do not burn out).

[0068] b) Voltage limit ellipse: This is precisely the core of the invention: under the high voltage of the 1612V bus, the system sets a safe phase voltage "ceiling" for the 380V motor. Due to AC / DC inductance , With a back electromotive force of up to 15mH and extremely high speed, this elliptical constraint forced system automatically reduces the current in the weak magnetic region to prevent high voltage breakdown.

[0069] 2. High-speed bilinear interpolation: 2D-LUT with output coefficient Using the normalized measured speed signal as coordinates, a bilinear interpolation algorithm is employed to instantly look up a table and output the electromagnetic torque command (following the maximum torque-to-current ratio (MTPA) criterion below the base speed). Subsequently, the gain is used to convert it into the quadrature-axis command current. This feedforward lookup method completely bypasses the real-time calculation of complex equations, achieving a response speed in the microsecond range.

[0070] IV. Bottom-level current closed-loop and limiting drive Cross-axis command current With direct axis command current Subtract each of its corresponding feedback current components to obtain shaft and Shaft current error signal; shaft and The current error signals of the shafts are fed into corresponding discrete-time PID controllers for adjustment. To address the high-frequency transients introduced by the LC filter, this embodiment employs a low proportional gain and medium integral gain (e.g., ...) for the current loop. After robust tuning, the calculated voltage command is passed through the insulation withstand boundary (380V standard) for secondary amplitude limiting, and then sent to the SVPWM (Space Vector Pulse Modulation) generator to finally generate the PWM signal to drive the inverter.

[0071] V. Dynamic Performance Demonstration Under the aforementioned hardware-software co-engineering architecture, this embodiment demonstrates dominant performance under harsh operating conditions, such as... Figure 5 As shown: High Inertia Soft Start: Facing The system, with its immense dead weight, was given a target speed of 3000 RPM. The fuzzy engine and 2D-LUT instantly released the maximum feedforward torque, and the system rapidly climbed to 3000 RPM in only about 1200 ms, with the speed overshoot strictly maintained at 0% throughout the entire process.

[0072] Heavy-load sudden-applied anti-jamming: During steady-state cruising at 3000 RPM (where the speed fluctuation error is within ±0.5 RPM), a sudden application of 10 Nm of heavy load is applied. The maximum instantaneous drop in system speed is extremely small (<5 RPM). The cascaded fuzzy algorithm instantly generates torque to smooth out the disturbance in less than 200 ms, demonstrating unparalleled heavy-load disturbance rejection and rapid correction capabilities.

[0073] The electronic device of this invention includes a central processing unit (CPU), which can perform various appropriate actions and processes according to computer program instructions stored in read-only memory (ROM) or loaded from a storage unit into random access memory (RAM). The RAM may also store various programs and data required for device operation. The CPU, ROM, and RAM are interconnected via a bus. Input / output (I / O) interfaces are also connected to the bus.

[0074] Multiple components in the device are connected to the I / O interface, including: input units such as keyboards and mice; output units such as various types of displays and speakers; storage units such as disks and optical discs; and communication units such as network interface cards (NICs), modems, and wireless transceivers. The communication unit allows the device to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0075] The processing unit performs the various methods and processes described above. For example, in some embodiments, the methods may be implemented as computer software programs tangibly contained in a machine-readable medium, such as a storage unit. In some embodiments, part or all of the computer program may be loaded and / or installed on the device via ROM and / or a communication unit. When the computer program is loaded into RAM and executed by the CPU, one or more steps of the methods described above may be performed. Alternatively, in other embodiments, the CPU may be configured to execute the methods by any other suitable means (e.g., by means of firmware).

[0076] The functions described above in this document can be performed, at least in part, by one or more hardware logic components. For example, exemplary types of hardware logic components that can be used, without limitation, include: Field Programmable Gate Arrays (FPGAs), Application-Specific Integrated Circuits (ASICs), Application Standard Products (ASSPs), System-on-Chip (SoCs), Complex Programmable Logic Devices (CPLDs), and so on.

[0077] The program code used to implement the methods of the present invention can be written in any combination of one or more programming languages. This program code can be provided to a processor or controller of a general-purpose computer, special-purpose computer, or other programmable data processing device, such that when executed by the processor or controller, the program code causes the functions / operations specified in the flowcharts and / or block diagrams to be implemented. The program code can be executed entirely on the machine, partially on the machine, as a standalone software package partially on the machine and partially on a remote machine, or entirely on a remote machine or server.

[0078] In the context of this invention, a machine-readable medium can be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. Machine-readable media can include, but are not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory, optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0079] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in the present invention, and these modifications or substitutions should all be covered within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A cascaded fuzzy feedforward direct drive speed control method based on LC filtering, characterized in that, The method includes: The measured speed signal of the low-voltage permanent magnet synchronous motor is acquired in real time, and the speed error and error change rate are obtained from the reference speed and the measured speed. The speed error and the rate of change of error are input into the Mamdani fuzzy inference engine, and the output force given coefficient is output. The output given coefficient and the normalized measured speed signal are input into a two-dimensional torque-current reference lookup table, bilinear interpolation calculation is performed, and an electromagnetic torque command is output; the two-dimensional torque-current reference lookup table has a built-in bilinear interpolation algorithm and discrete electromagnetic torque grid data; The electromagnetic torque command is converted into quadrature axis command current, which is then used to generate a PWM signal through current closed-loop regulation and SVPWM modulation to drive a two-level voltage-type inverter. The output of the two-level voltage source inverter drives a low-voltage permanent magnet synchronous motor via an LC sine wave filter with an RC damping branch.

2. The cascaded fuzzy feedforward direct drive speed control method based on LC filtering according to claim 1, characterized in that, The process of obtaining the output given coefficient includes: The speed error and the rate of change of error are multiplied by the quantization gain and mapped to the fundamental universe of discourse. Fuzzy inference is performed based on preset fuzzy sets and inference rules; The fuzzy inference results are defuzzified and amplitude-limited to obtain the output given coefficient.

3. The cascaded fuzzy feedforward direct drive speed control method based on LC filtering according to claim 2, characterized in that, The fuzzy inference based on preset fuzzy sets and inference rules specifically refers to: Within the basic domain, the rotational speed error and the rate of change of error are both divided based on a preset fuzzy set; the preset fuzzy set is {NB, NM, NS, ZE, PS, PM, PB}. An asymmetric rule matrix is ​​formed by the intersection of fuzzy sets of speed error and error change rate. Its boundary mapping conditions include: when the input speed error is NB and the error change rate is NB, the fuzzy set of the mapping output is NS; when the input speed error is PB and the error change rate is PB, the fuzzy set of the mapping output is PB.

4. The cascaded fuzzy feedforward direct drive speed control method based on LC filtering according to claim 1, characterized in that, The discrete electromagnetic torque grid data simultaneously satisfies the current-limiting circle constraint and the voltage-limiting ellipse constraint. The discrete electromagnetic torque grid data corresponds to the analytical solution of the maximum torque-to-current ratio control criterion in the constant torque region below the base speed, and to the analytical solution of the torque maximization control criterion under voltage constraints in the weak magnetic region above the base speed.

5. The cascaded fuzzy feedforward direct drive speed control method based on LC filtering according to claim 1, characterized in that, Using the output given coefficient and the normalized measured speed signal as coordinate input values, bilinear interpolation calculation is performed based on the bilinear interpolation algorithm in the two-dimensional torque current reference lookup table to output the electromagnetic torque command; the electromagnetic torque command is converted into the quadrature axis command current through torque-current gain conversion. Simultaneously, the normalized measured speed signal is input into a one-dimensional torque current reference lookup table, and the direct-axis command current for field weakening control is output after the lookup table is consulted.

6. The cascaded fuzzy feedforward direct drive speed control method based on LC filtering according to claim 5, characterized in that, The PWM signal generated by current closed-loop regulation and SVPWM modulation includes: The feedback current of the low-voltage permanent magnet synchronous motor is obtained. The difference between the quadrature axis command current, the direct axis command current and the corresponding feedback current is calculated. The two difference calculation results are input into the corresponding discrete-time PID controllers, and after independent adjustment, AC voltage command and DC voltage command are output respectively. The AC voltage command and DC voltage command are amplitude-limited and then input into the SVPWM generator, where they are modulated to output a PWM signal.

7. The cascaded fuzzy feedforward direct drive speed control method based on LC filtering according to claim 1, characterized in that, The LC sine wave filter with RC damping branch includes a three-phase filter inductor, a three-phase filter capacitor, and an RC damping branch. The three-phase filter inductor is connected in series between the two-level voltage type inverter and the low-voltage permanent magnet synchronous motor; The three-phase filter capacitors are connected in parallel between the phases on the output side of the three-phase filter inductor; The RC damping branch is connected in parallel with the three-phase filter capacitor, and within the RC damping branch, the damping resistor and the damping capacitor are connected in series.

8. A cascaded fuzzy feedforward direct-drive speed control system based on LC filtering, characterized in that, This system is used to implement a cascaded fuzzy feedforward direct drive speed control method based on LC filtering as described in any one of claims 1 to 7, wherein the system includes a main power circuit link and a cascaded control system; The main power circuit link includes a two-level voltage source inverter and an LC sine wave filter with an RC damping branch. The cascaded control system includes a subtractor, a cascaded fuzzy decision module, a dynamic mapping execution module, a current command calculation module, and a current closed-loop and limiting drive module connected in sequence. The reference speed and the real-time measured speed signal distribution of the low-voltage permanent magnet synchronous motor are input to the two input terminals of the subtractor, and the output speed error and error change rate are output. The speed error and the rate of change of error are input into the cascaded fuzzy decision module for Mamdani fuzzy inference, and the output force given coefficient is output. The output given coefficient and the normalized measured speed signal are input into the power mapping execution module, which performs bilinear interpolation calculation based on a two-dimensional torque-current reference lookup table and outputs an electromagnetic torque command; the two-dimensional torque-current reference lookup table has a built-in bilinear interpolation algorithm and discrete electromagnetic torque grid data; The current command calculation module converts the received electromagnetic torque command into quadrature axis command current. The quadrature axis command current is then adjusted by the current closed loop and the limiting drive module, and SVPWM modulation is used to generate a PWM signal to drive the two-level voltage type inverter. The output of the two-level voltage source inverter drives a low-voltage permanent magnet synchronous motor via an LC sine wave filter with an RC damping branch.

9. A cascaded fuzzy feedforward direct-drive speed control system based on LC filtering according to claim 8, characterized in that, The power mapping execution module includes a two-dimensional torque-current reference lookup table and a one-dimensional torque-current reference lookup table. The two input terminals of the two-dimensional torque current reference lookup table are respectively connected to the output terminal of the cascaded fuzzy decision module and the normalization terminal of the measured speed signal, and the output terminal is connected to the input terminal of the current command calculation module. The output of the current command calculation module is connected to the current closed-loop and limiting drive module. The input of the one-dimensional torque current reference lookup table is connected to the normalized result of the measured speed signal, and the output is connected to the current closed-loop and limiting drive module.

10. A cascaded fuzzy feedforward direct-drive speed control system based on LC filtering according to claim 8, characterized in that, The two-dimensional torque-current reference lookup table incorporates a bilinear interpolation algorithm and discrete electromagnetic torque grid data, wherein the discrete electromagnetic torque grid data simultaneously satisfies the current-limiting circle constraint and the voltage-limiting ellipse constraint.

Images