Magnetic suspension motor control method and device and magnetic suspension motor

Abstract

The invention provides a magnetic suspension motor control method and device and a magnetic suspension motor, and relates to the technical field of magnetic suspension motor control, and the technical scheme is characterized in that the method comprises the steps: detecting the voltage change rate of a power grid, and judging a voltage fluctuation mode based on a preset threshold value; environment RFID tag information is read through a scene code analysis unit, and a current scene code is obtained; retrieving a maximum compensation current value, a current change rate threshold value and a PID integral time constant from a preset parameter matrix based on the voltage fluctuation mode and the scene code; and limiting the current change rate of the motor winding according to the current change rate threshold, and executing motor control through the maximum compensation current value and the PID integral time constant. The magnetic suspension motor control method and device and the magnetic suspension motor provided by the invention have the advantages of improving the operation stability of the motor in voltage fluctuation and multi-noise scenes, suppressing electromagnetic noise and adapting to different scenes.

Description

Technical Field

[0001] The present application relates to the technical field of magnetic levitation motor control, and in particular to a magnetic levitation motor control method, device and magnetic levitation motor. Background Art

[0002] During the operation of low-power distilled water machines, grid load fluctuations and multi-scenario noise control requirements pose dual challenges to the stability and quietness of magnetic levitation motors. The traditional control method uses a fixed-parameter PID strategy. Although it can maintain basic operation when the grid voltage is stable, it has significant defects when the voltage fluctuates violently. Since the low-power motor winding adopts a low-resistance design to improve energy efficiency, its reduced inductance value leads to a significant increase in the current change rate, and a small voltage fluctuation can cause a sudden change in current. This mutation not only aggravates electromagnetic noise, but is also likely to cause motor instability due to current overshoot. Especially in medical or laboratory scenarios with extremely high noise sensitivity requirements, traditional methods cannot simultaneously meet the needs of stable equipment operation and dynamic noise suppression. In the prior art, in order to cope with voltage sag, a compensation method of increasing the driving current is often used. However, this method causes the temperature rise of the motor winding to accelerate, which in turn aggravates the stator resistance drift, forming a vicious cycle of parameter mismatch. On the other hand, although simply limiting the current change rate can reduce noise, it cannot adapt to the differences in noise thresholds in different scenarios. For example, the acceptable medium noise level in a home scenario may exceed the allowable range in a medical scenario, and fixed parameter control lacks the ability to identify and adaptively adjust the operating scenario. In addition, the traditional method does not consider the impact of load conditions (such as water level and water temperature) on motor parameter drift, resulting in a mismatch between compensation parameters and real-time operating conditions, further weakening control accuracy. These limitations make it difficult for low-power magnetic levitation motors to balance energy efficiency, stability and quietness in complex power grid environments and multi-scenario applications. In view of the above problems, the existing technology needs to be improved urgently. Summary of the invention

[0003] The purpose of the present application is to provide a magnetic levitation motor control method, device and magnetic levitation motor, which have the advantages of improving the operating stability of the motor under voltage fluctuations and high-noise scenarios, suppressing electromagnetic noise and adapting to the needs of different scenarios. The present application provides a magnetic levitation motor control method for controlling the operation of a magnetic levitation motor of a low-power distilled water machine in a scenario where the grid voltage fluctuation exceeds a threshold and multiple noise thresholds. The technical solution is as follows: Detecting the rate of change of grid voltage and determining the voltage fluctuation mode based on a preset threshold; The scene code parsing unit reads the environmental RFID tag information to obtain the current scene code; Based on the voltage fluctuation pattern and scenario coding, the maximum compensation current value, current change rate threshold and PID integration time constant are retrieved from the preset parameter matrix; The motor winding current change rate is limited according to the current change rate threshold, and the motor control is performed by the maximum compensation current value and the PID integral time constant. Furthermore, the present application also proposes that the step of retrieving the maximum compensation current value, the current change rate threshold and the PID integral time constant from the preset parameter matrix based on the voltage fluctuation mode and the scene coding includes: Collecting motor operation data in real time, the motor operation data including motor winding temperature, motor speed and stator current; Calculating the motor parameter drift based on the motor winding temperature, motor speed and stator current, wherein the motor parameters include stator resistance, mutual inductance and back electromotive force coefficient; According to the motor parameter drift, the least square method is used to update the preset parameter matrix, wherein the preset parameter matrix includes the maximum compensation current value, the current change rate threshold value and the PID integral time constant under different voltage fluctuation modes and scene coding combinations; In response to changes in the voltage fluctuation mode and the scene coding, the updated preset parameter matrix is ​​queried to obtain the corresponding maximum compensation current value, current change rate threshold and PID integral time constant. Furthermore, the present application also proposes that the step of updating the preset parameter matrix using the least square method according to the motor parameter drift, wherein the preset parameter matrix includes the maximum compensation current value, the current change rate threshold and the PID integral time constant under different voltage fluctuation modes and scene coding combinations includes: Obtaining the current load status of the distilled water machine, wherein the load status includes water level and water temperature; Based on the water level and the water temperature, determining a motor parameter drift weight coefficient corresponding to the current load state, wherein the motor parameter drift weight coefficient includes a stator resistance weight coefficient, a mutual inductance weight coefficient, and a back electromotive force weight coefficient; According to the motor parameter drift amount and the motor parameter drift weight coefficient, the weighted least squares method is used to update the preset parameter matrix, which includes the maximum compensation current value, the current change rate threshold and the PID integral time constant under different voltage fluctuation modes and scene coding combinations. Furthermore, the present application also proposes that the step of detecting the grid voltage change rate and determining the voltage fluctuation mode based on a preset threshold comprises: Collect grid voltage data in real time and filter it using a sliding average filtering algorithm; Calculate the change rate of the filtered voltage data within a preset time window to obtain the voltage change rate; Perform spectrum analysis on the voltage change rate and extract the main frequency component of voltage fluctuation; When the voltage change rate exceeds a preset threshold and the main frequency component is lower than a preset frequency threshold, the voltage fluctuation mode is determined. Furthermore, the present application also proposes that the step of updating the preset parameter matrix using the weighted least squares method according to the motor parameter drift amount and the motor parameter drift weight coefficient, wherein the preset parameter matrix includes the maximum compensation current value, the current change rate threshold and the PID integral time constant under different voltage fluctuation modes and scene coding combinations includes: Constructing an objective function, wherein the objective function takes the maximum compensation current value, the current change rate threshold value and the PID integral time constant in the preset parameter matrix as independent variables, takes the motor parameter drift amount and the motor parameter drift weight coefficient as constraint conditions, and the objective function is used to minimize the motor control error; A sequential quadratic programming algorithm is used to solve the objective function and obtain an updated preset parameter matrix, which includes a maximum compensation current value, a current change rate threshold and a PID integral time constant under different voltage fluctuation modes and scene coding combinations. Furthermore, the present application also proposes that the step of retrieving the maximum compensation current value, the current change rate threshold and the PID integral time constant from the preset parameter matrix based on the voltage fluctuation mode and the scene coding includes: Determine the voltage level corresponding to the voltage fluctuation pattern and the noise level corresponding to the scene coding; Based on the voltage level and the noise level, matching is performed in an index table of a preset parameter matrix, wherein the index table stores a mapping relationship between the voltage level, the noise level and the parameter storage address; In response to the presence of a parameter storage address in the index table that matches both the voltage level and the noise level, a maximum compensation current value, a current change rate threshold, and a PID integral time constant are read from the parameter storage address. Furthermore, the present application also proposes that the steps of limiting the rate of change of the motor winding current according to the current change rate threshold and performing motor control through the maximum compensation current value and the PID integral time constant include: Calculate the difference between the actual rate of change of the motor winding current and the current rate of change threshold; When the difference is greater than zero, the current gradient limiter is started to reduce the motor winding voltage by adjusting the PWM duty cycle, thereby suppressing the actual change rate of the motor winding current and making the actual change rate converge to the current change rate threshold; Based on the maximum compensation current value, the target current value of the PID controller is calculated. According to the target current value and the PID integral time constant, the output voltage of the PID controller is adjusted to compensate for the influence of voltage fluctuation on the motor operation. Furthermore, the present application also proposes that when the difference is greater than zero, the current gradient limiter is started, the motor winding voltage is reduced by adjusting the PWM duty cycle, and the actual change rate of the motor winding current is suppressed so that the actual change rate converges to the current change rate threshold, the steps include: When the difference is greater than zero, the current gradient limiter is started, and the PWM duty cycle adjustment rate is determined in response to the voltage fluctuation mode, the voltage sag mode corresponds to a fast adjustment rate, and the voltage slow change mode corresponds to a gentle adjustment rate; Based on the scene code, determine the PWM switching frequency; The PWM duty cycle is adjusted according to the PWM duty cycle adjustment rate, and the motor winding voltage is driven at the PWM switching frequency to suppress the actual change rate of the motor winding current so that the actual change rate converges to the current change rate threshold. Furthermore, the present application also proposes a magnetic levitation motor control device for controlling the operation of a magnetic levitation motor of a low-power distilled water machine in a scenario where the grid voltage fluctuation exceeds a threshold and multiple noise thresholds, comprising: A voltage detection unit, used to detect the rate of change of the grid voltage and determine the voltage fluctuation mode based on a preset threshold; A scene code parsing unit is used to read the environmental RFID tag information and obtain the current scene code; A parameter retrieval unit, used to retrieve a maximum compensation current value, a current change rate threshold, and a PID integral time constant from a preset parameter matrix based on a voltage fluctuation mode and a scenario code; The control execution unit is used to limit the rate of change of the motor winding current according to the current change rate threshold value, and to perform motor control through the maximum compensation current value and the PID integral time constant. Furthermore, the present application also proposes a magnetic levitation motor, to which the above method is applied. From the above, it can be seen that the magnetic levitation motor control method, device and magnetic levitation motor provided in the present application detect the grid voltage fluctuation pattern and dynamically adjust the control parameters in combination with scene coding, limit the current change rate and adaptively compensate for voltage fluctuations, effectively suppress electromagnetic noise and improve motor stability, and have the beneficial effects of adapting to multiple scenario requirements, improving operation accuracy and reducing the risk of parameter mismatch. BRIEF DESCRIPTION OF THE DRAWINGS

[0004] Figure 1 A schematic flow chart of a magnetic levitation motor control method provided in this application.

[0005] Figure 2 A schematic structural diagram of a magnetic levitation motor control device provided in this application.

[0006] In the figure: 210, voltage detection unit; 220, scene coding analysis unit; 230, parameter retrieval unit; 240, control execution unit. DETAILED DESCRIPTION

[0007] The technical solutions in the present application will be clearly and completely described below in conjunction with the drawings in the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. The components of the present application generally described and shown in the drawings here can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the present application for protection, but merely represents the selected embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without making creative work belong to the scope of protection of the present application.

[0008] It should be noted that similar reference numerals and letters represent similar items in the following drawings, so once an item is defined in one drawing, it does not need to be further defined and explained in subsequent drawings. At the same time, in the description of this application, the terms "first", "second", etc. are only used to distinguish the description and cannot be understood as indicating or implying relative importance.

[0009] In the control of the magnetic levitation motor of the traditional distilled water machine, the PID control strategy with fixed parameters cannot adapt to the dynamic coupling of grid voltage fluctuations and multi-scenario noise constraints. When the grid voltage fluctuation amplitude exceeds the threshold, the voltage change rate causes a sudden change in the motor winding current, and the low-resistance winding design further amplifies the current change rate. At the same time, there are significant differences in the threshold limits of electromagnetic noise in scenarios such as medical areas and laboratories, and fixed control parameters are difficult to meet the dual requirements of rapid current compensation and noise suppression at the same time. The separation of voltage fluctuation detection and scene noise constraints causes the motor to be forced to choose a suboptimal control state of current overshoot or noise exceeding the standard when the voltage drops suddenly.

[0010] For example, when the medical area is operating during the off-peak period of the power grid load at night, multiple high-power devices are shut down at the same time, causing a sudden rise in the local power grid voltage. At this time, the magnetic levitation motor of the distilled water machine needs to quickly increase the driving current to maintain suspension stability, but the threshold limit of electromagnetic noise in the medical environment requires the current change rate to be lower than that in normal scenarios. The traditional control method does not associate the scene encoding when retrieving the parameter matrix, and only selects the compensation current based on the voltage fluctuation mode, causing the current gradient to exceed the noise threshold of the medical area and trigger the equipment alarm. If the current change rate is forcibly limited, the motor will have axial displacement deviation due to insufficient compensation, and eventually trigger a protective shutdown.

[0011] In this regard, refer to Figure 1The present application proposes a magnetic levitation motor control method for controlling the operation of a magnetic levitation motor of a low-power distilled water machine in a scenario where the grid voltage fluctuation exceeds a threshold and multiple noise thresholds, comprising the following steps: S110, detecting a grid voltage change rate, and determining a voltage fluctuation mode based on a preset threshold; S120, reading the environmental RFID tag information through the scene code parsing unit to obtain the current scene code; S130, based on the voltage fluctuation mode and the scene coding, retrieving the maximum compensation current value, the current change rate threshold value and the PID integral time constant from the preset parameter matrix; S140, limiting the rate of change of the motor winding current according to the current change rate threshold, and performing motor control through the maximum compensation current value and the PID integral time constant.

[0012] Among them, grid voltage change rate detection refers to identifying abnormal voltage conditions by real-time monitoring of the grid voltage fluctuation speed. Specifically, it can be achieved by using a differential voltage sensor in conjunction with a sliding window average algorithm to determine voltage sag or slow change mode, thereby triggering different control strategies. Among them, the scene coding and parsing unit refers to the acquisition of the noise level identification of the environment in which the equipment is located through radio frequency identification technology. Specifically, it can be implemented using a UHF band RFID reader and an encrypted tag communication protocol, and is used to dynamically identify the noise threshold limits of different scenarios such as medical areas and laboratories. Among them, the preset parameter matrix refers to a database structure for storing multi-dimensional control parameters, which can be implemented by using a three-dimensional array to establish an index table according to voltage mode, scene coding, and load status, so as to quickly match the optimal control parameter combination under the current working conditions. Among them, the current change rate threshold limit refers to the dynamic constraint on the rising rate of the motor drive current, which can be achieved by collaborative control of the gradient observer and the PWM duty cycle closed-loop regulation to suppress the excessive electromagnetic noise caused by current mutations. Among them, PID integral time constant adjustment refers to dynamically correcting the integral response speed of the controller according to the voltage fluctuation characteristics. It can be achieved by combining fuzzy logic with an online parameter tuning algorithm to balance the contradiction between voltage compensation speed and system stability. The core innovation of this application is to establish a dual matching mechanism of voltage fluctuation pattern and scene coding, and realize dynamic coupling adjustment of control parameters through a preset parameter matrix, so as to maintain the stable operation of the motor under grid fluctuations while suppressing excessive noise caused by current mutation.

[0013] The working process and principle of this application is to first detect the grid voltage change rate and determine the voltage fluctuation mode through a preset threshold. The voltage detection unit continuously samples the grid voltage and calculates the voltage change rate between adjacent sampling points. When the change rate exceeds the preset threshold, the voltage fluctuation mode determination is triggered. The voltage fluctuation modes include voltage sag, voltage swell, and voltage slow change.

[0014] Secondly, the scene code parsing unit reads the environmental RFID tag information and obtains the current scene code. The RFID reader scans the preset RFID tags in the environment and decodes them to obtain the scene code. The scene code corresponds to different application environments, such as medical areas, laboratories, office areas, etc. Each scene has a specific noise threshold requirement.

[0015] Furthermore, based on the voltage fluctuation mode and scenario coding, the maximum compensation current value, current change rate threshold and PID integral time constant are retrieved from the preset parameter matrix. The parameter matrix pre-stores the optimal control parameters under different voltage fluctuation modes and scenario coding combinations. Through the combined index of voltage fluctuation mode and scenario coding, the control parameters suitable for the current working conditions can be quickly retrieved.

[0016] Finally, the motor winding current change rate is limited according to the current change rate threshold, and the motor control is performed through the maximum compensation current value and the PID integral time constant. The current gradient limiter constrains the current change speed according to the current change rate threshold and suppresses electromagnetic noise. The PID controller sets the target current based on the maximum compensation current value and optimizes the dynamic response characteristics by adjusting the integral time constant.

[0017] Therefore, the present application realizes the dynamic adaptation of control parameters with voltage fluctuations and scene requirements, overcoming the limitations of traditional fixed parameter control. Through the dual judgment mechanism of voltage fluctuation detection and scene coding recognition, the coordinated optimization of motor control and noise suppression is realized.

[0018] As a preferred embodiment, the solution of the present application is specifically implemented as follows: The voltage detection unit uses a high-precision voltage sampling chip. The voltage change rate threshold is set to 5V / ms, which is used to trigger the voltage fluctuation mode judgment. The voltage sag mode corresponds to a change rate less than -5V / ms, the voltage swell mode corresponds to a change rate greater than 5V / ms, and the voltage slow change mode corresponds to a change rate between ±5V / ms.

[0019] The scene code parsing unit uses an RFID card reader module. RFID tags are preset in the environment, for example, the medical area is coded as 0x01, the laboratory is coded as 0x02, and the office area is coded as 0x03. The RFID reader scans the environment tag every 5 seconds to update the scene code.

[0020] The parameter matrix is ​​stored in a two-dimensional array, with the row index being the voltage fluctuation mode and the column index being the scene code. The matrix elements contain the maximum compensation current value, the current change rate threshold, and the PID integral time constant. For example, the parameter combination of the medical area in the voltage sag mode is {2A, 0.5A / ms, 10ms}, and the parameter combination of the laboratory is {2.5A, 1A / ms, 8ms}.

[0021] The current gradient limiter controls the current change rate by adjusting the PWM duty cycle. When the actual current change rate exceeds the threshold, the limiter linearly reduces the PWM duty cycle until the current change rate converges to the threshold.

[0022] The PID controller uses an incremental PID algorithm. The integral time constant can be dynamically adjusted within the range of 5ms-20ms to optimize the dynamic response characteristics of the motor. The maximum compensation current value is used as the output limiter of the PID controller to prevent current oscillation caused by over-regulation.

[0023] Through the above scheme, this application realizes the stable operation of the magnetic levitation motor of the low-power distilled water machine under the scenarios of grid voltage fluctuations and multiple noise thresholds. The collaborative mechanism of voltage fluctuation detection and scene coding recognition enables the control parameters to quickly adapt to the grid status and environmental requirements. The combination of current gradient limitation and dynamic adjustment of PID parameters effectively suppresses electromagnetic noise while ensuring the stability of the motor. This solution overcomes the limitations of traditional fixed parameter control under complex working conditions and improves the adaptability and reliability of low-power distilled water machines in noise-sensitive scenarios such as medical and laboratories.

[0024] In some of the above-mentioned schemes of the present application, a method of retrieving control parameters from a preset parameter matrix based on voltage fluctuation patterns and scene coding is proposed to realize dynamic control of the motor. However, due to the actual changes in parameters such as winding temperature, speed, stator current, etc. during the operation of the motor, key parameters such as the motor stator resistance, mutual inductance and back electromotive force coefficient will drift. If the initial preset parameter matrix is ​​still used directly, the retrieved maximum compensation current value, current change rate threshold and PID integral time constant will not match the actual operating conditions, thereby causing problems such as control error accumulation, decreased motor operation stability and noise suppression failure.

[0025] In this regard, the present application further proposes: real-time collection of motor operation data, the motor operation data including motor winding temperature, motor speed and stator current; based on the motor winding temperature, motor speed and stator current, calculating the motor parameter drift, the motor parameters including stator resistance, mutual inductance and back electromotive force coefficient; according to the motor parameter drift, using the least squares method to update the preset parameter matrix, the preset parameter matrix including the maximum compensation current value, current change rate threshold and PID integral time constant under different voltage fluctuation modes and scene coding combinations; in response to changes in voltage fluctuation modes and scene coding, query the updated preset parameter matrix to obtain the corresponding maximum compensation current value, current change rate threshold and PID integral time constant.

[0026] Among them, the temperature of the motor winding can be collected by the NTC temperature sensor buried in the stator slot; the motor speed is measured by the Hall sensor or the photoelectric encoder; the stator current is collected by the current transformer, and its range needs to cover 120% of the rated current of the motor. The parameter drift calculation adopts the Kalman filter algorithm. By establishing the observation equation containing the temperature-resistance transfer function and the speed-back electromotive force coefficient dynamic model, the motor winding temperature is used as the main variable of the resistance drift, and the motor speed is used as the correction factor of the back electromotive force coefficient drift. The preset parameter matrix adopts a two-dimensional index structure. The row vector corresponds to the voltage fluctuation mode classification, the column vector corresponds to the scene coding classification, and the matrix elements store the control parameter combination after normalization. In the least squares update process, the parameter drift is converted into the residual term of the parameter matrix, and the objective function is converged to the allowable error range through iterative calculation.

[0027] Specifically, the motor winding temperature, speed and stator current data are synchronously input into the parameter drift estimation module. The temperature data is processed by sliding average filtering and then substituted into the stator resistance temperature coefficient formula to calculate the real-time resistance value. The deviation from the initial resistance value is used as the resistance drift. The speed data is used to correct the speed-voltage relationship in the back electromotive force coefficient model, and the parameter coupling error caused by speed fluctuation is eliminated through the dynamic compensation algorithm. During the parameter matrix update process, the least squares method uses the current parameter drift as a constraint condition to perform linear regression analysis on each control parameter in the matrix. For example, a negative correlation is established between the maximum compensation current value and the stator resistance drift, and a positive correlation is established between the current change rate threshold and the back electromotive force coefficient drift. The updated parameter matrix is ​​optimized for storage through a hash table, and the voltage fluctuation mode and scene coding combination are converted into a hash key value to achieve fast retrieval of control parameters. When a voltage fluctuation mode or scene coding change is detected, the updated parameter matrix is ​​reloaded into the control decision module to ensure that the PID integral time constant matches the current motor electromagnetic characteristics, thereby suppressing current oscillation and electromagnetic noise caused by parameter mismatch.

[0028] Through the above technical scheme, the present application realizes the dynamic compensation of the influence of the drift of stator resistance, mutual inductance and back-electromotive force coefficient on the control parameters during the operation of the motor, and optimizes the preset parameter matrix online to make the maximum compensation current value, current change rate threshold and PID integral time constant match the actual working condition of the motor in real time, effectively solving the current loop oscillation and torque pulsation problems caused by parameter mismatch, ensuring that the magnetic levitation motor maintains stable suspension and reduces high-frequency electromagnetic noise under grid voltage fluctuations and high-noise scenarios.

[0029] In some of the above-mentioned schemes of the present application, it is proposed to use the least square method to update the preset parameter matrix according to the motor parameter drift to dynamically adjust the control parameters. However, in this process, the difference in the impact of different load states of the distilled water machine on the motor parameter drift is not considered. Since there are significant differences in the temperature rise gradient and current dynamic response characteristics of the motor winding when the low-power distilled water machine is running at low load and high load, if the same weight is used for the parameter drift under all working conditions, the parameter matrix update will deviate from the actual operating state, thereby reducing the motor control accuracy and noise suppression effect.

[0030] In this regard, the present application further proposes to obtain the current load state of the distilled water machine, wherein the load state includes water level and water temperature; based on the water level and water temperature, determine the motor parameter drift weight coefficient corresponding to the current load state, wherein the motor parameter drift weight coefficient includes the stator resistance weight coefficient, the mutual inductance weight coefficient and the back electromotive force coefficient weight coefficient; according to the motor parameter drift amount and the motor parameter drift weight coefficient, use the weighted least squares method to update the preset parameter matrix, wherein the preset parameter matrix includes the maximum compensation current value, the current change rate threshold and the PID integral time constant under different voltage fluctuation modes and scene coding combinations.

[0031] Among them, the acquisition of load state is realized by the water level sensor and temperature sensor installed inside the water tank, and the water level sensor adopts the capacitive measurement principle. The determination of the motor parameter drift weight coefficient is realized by a preset two-dimensional mapping table, which takes the water level percentage and the water temperature in Celsius as input, and outputs the stator resistance weight coefficient, the mutual inductance weight coefficient and the back electromotive force coefficient weight coefficient. As a preferred embodiment, the water level is divided into three gears: low (<30%), medium (30%-70%), and high (>70%), and the water temperature is divided into three gears: low temperature (<60℃), medium temperature (60℃-80℃), and high temperature (>80℃). The corresponding stator resistance weight coefficient ranges from 0.5-1.2, the mutual inductance weight coefficient is 0.8-1.5, and the back electromotive force coefficient weight coefficient is 1.0-2.0. The application of weighted least squares method requires the construction of a weighted residual sum of squares function, whose weight matrix consists of three weight coefficients forming the diagonal matrix elements. The specific form is W=diag(α_R,α_L,α_ψ), where α_R is the stator resistance weight coefficient, α_L is the mutual inductance weight coefficient, and α_ψ is the back electromotive force coefficient weight coefficient.

[0032] Specifically, when the grid voltage fluctuation is detected during the operation of the magnetic levitation motor, the current load state parameters of the distilled water machine are first continuously collected through the water level sensor and the temperature sensor. The measured water level value and water temperature value are input into the preset two-dimensional mapping table, and the corresponding three weight coefficients are queried. Then the real-time calculated stator resistance drift ΔR, mutual inductance drift ΔL and back electromotive force coefficient drift Δψ are respectively multiplied by the corresponding weight coefficients to form a weighted drift. In the parameter matrix update process, by constructing the objective function of the weighted least squares method, the various control parameters in the updated parameter matrix can give priority to reflecting the parameter drift that has a significant impact on the current load state. For example, under high temperature and high water level conditions, the back electromotive force coefficient weight coefficient is set to 2.0. At this time, the influence weight of Δψ in the parameter matrix update is doubled, thereby more accurately compensating for the demagnetization effect of the permanent magnet caused by high temperature. This process is combined with voltage fluctuation pattern recognition and scene coding retrieval, so that the maximum compensation current value, current change rate threshold and PID integral time constant finally obtained can match the current grid state and adapt to the actual load conditions, thereby achieving improved motor control accuracy and effective noise suppression.

[0033] As a preferred embodiment, the solution of the present application is specifically implemented as follows: Get the current load status of the distilled water machine, including water level and water temperature. The water level is detected by the water level sensor, and the water temperature is detected by the temperature sensor. For example, the water level is 80% and the water temperature is 60°C.

[0034] Based on the water level and water temperature, determine the motor parameter drift weight coefficient corresponding to the current load state. Specifically, establish a mapping table of load state and motor parameter drift weight coefficient. When the water level is 80% and the water temperature is 60°C, the stator resistance weight coefficient is 0.8, the mutual inductance weight coefficient is 0.6, and the back electromotive force coefficient weight coefficient is 0.7.

[0035] According to the motor parameter drift and the motor parameter drift weight coefficient, the weighted least square method is used to update the preset parameter matrix. Among them, the motor parameter drift is calculated by the motor operation data collected in real time. For example, the stator resistance drift is 5%, the mutual inductance drift is 3%, and the back electromotive force coefficient drift is 4%. Multiply the motor parameter drift with the corresponding weight coefficient to obtain the weighted drift. Then construct the least squares objective function, with the maximum compensation current value, current change rate threshold and PID integral time constant in the preset parameter matrix as independent variables, and the weighted drift as the constraint condition, solve the objective function to obtain the updated preset parameter matrix.

[0036] Through the above technical solution, the present application can dynamically adjust the weight of the motor parameter drift according to the actual load state of the distilled water machine, and improve the accuracy and pertinence of the parameter matrix update. As a result, under different load conditions, the control system can better adapt to the actual drift of the motor parameters, improve the motor control accuracy, and effectively suppress noise. By considering the impact of the load state on the motor parameter drift, this solution avoids the problem of the parameter matrix update deviating from the actual operating state due to the use of a unified weight to handle parameter drift under all working conditions, thereby improving the adaptability and stability of low-power distilled water machines in complex power grid environments and multiple noise threshold scenarios.

[0037] In some of the above-mentioned schemes of the present application, when collecting grid voltage data in real time, due to the presence of high-frequency noise and instantaneous interference in the grid signal, directly calculating the voltage change rate will lead to misjudgment of the voltage fluctuation pattern. At the same time, traditional detection methods cannot distinguish between rapid voltage fluctuations and continuous low-frequency disturbances, resulting in inaccurate control parameter matching.

[0038] In this regard, the present application further proposes the steps of detecting the rate of change of grid voltage and determining the voltage fluctuation mode based on a preset threshold, including: collecting grid voltage data in real time, filtering the grid voltage data using a sliding average filtering algorithm; calculating the rate of change of the filtered voltage data within a preset time window to obtain the voltage change rate; performing spectral analysis on the voltage change rate to extract the main frequency component of the voltage fluctuation; when the voltage change rate exceeds the preset threshold and the main frequency component is lower than the preset frequency threshold, determining the voltage fluctuation mode.

[0039] Among them, the sliding average filtering algorithm performs moving average processing on the grid voltage data by setting a sliding window with a length of 5-15 sampling cycles. The preset time window is set to the interval of 20-50ms, which can balance the capture of voltage fluctuation trends and the requirements of data stability. The spectrum analysis can use the wavelet transform algorithm to extract the amplitude of the main frequency component in the 0.1-5Hz frequency band, which covers the characteristic frequency of typical low-frequency disturbances in the power grid. The judgment condition that the main frequency component is lower than the preset frequency threshold forms a joint constraint with the time domain criterion of the voltage change rate exceeding the limit.

[0040] Specifically, after the grid voltage data is filtered by sliding average, the high-frequency noise components are filtered out, and the retained voltage signal fluctuation trend is calculated by the preset time window. After obtaining the voltage change rate, the main frequency component of the fluctuation is separated by spectrum analysis. When the main frequency component is in the low-frequency range and the change rate exceeds the preset threshold, it is judged as a continuous low-frequency disturbance mode of the grid. For example, when the voltage change rate is detected to be 5V / s and the main frequency component is 3Hz, it is compared with the preset threshold of 3V / s and the frequency threshold of 10Hz. The corresponding control strategy is triggered when the dual conditions are met. This combined judgment method effectively distinguishes the high-frequency transient characteristics of voltage sag events from the low-frequency continuous characteristics of low-frequency oscillations through time-frequency joint analysis, avoiding the risk of misjudgment under a single threshold judgment.

[0041] Through the above technical solution, the present application realizes accurate identification of grid voltage fluctuations. The sliding average filter effectively eliminates high-frequency noise and instantaneous interference, and retains the real voltage fluctuation trend. The setting of the preset time window balances the response speed and data stability. Spectral analysis analyzes the fluctuation characteristics from the dual dimensions of time domain and frequency domain, and distinguishes between sudden voltage drops and low-frequency oscillations of the grid. The combined judgment logic overcomes the defect that the traditional single threshold judgment is prone to misjudgment, provides reliable input for subsequent parameter matrix retrieval, and ensures the accuracy of dynamic adjustment of control parameters.

[0042] In some of the above-mentioned schemes of the present application, a scheme is proposed to update the preset parameter matrix by using the weighted least squares method combined with the load state weight coefficient. However, under the constraints of dynamic grid fluctuations and multi-scenario noise, it is difficult to effectively balance the coupling relationship between different control parameters by only linearly weighting the parameter drift through the weight coefficient, which can easily cause the objective function to converge to the local optimal solution, thereby causing insufficient update accuracy of the preset parameter matrix and affecting the global minimization of the motor control error.

[0043] In this regard, the present application further proposes: constructing an objective function, the objective function takes the maximum compensation current value, the current change rate threshold and the PID integral time constant in the preset parameter matrix as independent variables, and takes the motor parameter drift and the motor parameter drift weight coefficient as constraints, and the objective function is used to minimize the motor control error; using a sequential quadratic programming algorithm to solve the objective function and obtain an updated preset parameter matrix, the preset parameter matrix including the maximum compensation current value, the current change rate threshold and the PID integral time constant under different voltage fluctuation modes and scene coding combinations.

[0044] Among them, the construction of the objective function requires the maximum compensation current value, the current change rate threshold and the PID integral time constant as multi-dimensional optimization variables, and the physical change range of the parameters is constrained by the motor parameter drift. For example, the stator resistance drift is limited to ±5% of the nominal value, and the back electromotive force coefficient drift is subject to ±3% of the nominal value. At the same time, the motor parameter drift weight coefficient is introduced as a Lagrange multiplier to introduce constraints. In the specific implementation, the stator resistance weight coefficient ranges from 0.6 to 1.2, and the back electromotive force coefficient weight coefficient ranges from 0.8 to 1.5. Its value is dynamically adjusted according to the load state. The sequential quadratic programming algorithm approaches the optimal solution by iteratively solving the quadratic subproblem. Each iteration needs to calculate the Hessian matrix and gradient vector of the objective function, and combine the Jacobian matrix of the constraint condition to search for feasible directions, and finally converge to the global optimal solution within 3-5 iterations.

[0045] Specifically, under the constraints of grid voltage fluctuations and multi-scenario noise, the control parameters in the preset parameter matrix are first defined as optimization variables, and the inequality constraint boundaries are established through the motor parameter drift. For example, the maximum compensation current value must not exceed 120% of the rated current. Then the weight coefficient corresponding to the load state is used as the scaling factor of the constraint condition. For example, when the load water temperature is higher than 60°C, the weight coefficient of the back electromotive force coefficient is increased to 1.2 to enhance the temperature drift suppression. The objective function uses the root mean square value of the motor control error as the optimization index, and its mathematical expression is the integral square sum of the current tracking error under each working condition. When solving through the sequential quadratic programming algorithm, the nonlinear constraints need to be converted into quadratic programming sub-problems, and the active constraints are processed by the effective set method, and finally the updated parameter matrix that satisfies the global optimum is output. Therefore, the update process of the parameter matrix not only ensures the physical feasibility of the control parameters, but also overcomes the local optimal trap through nonlinear optimization, so that the motor control error can converge to the global minimum under multiple working conditions.

[0046] As a preferred embodiment, the solution of the present application is specifically implemented as follows: Construct the objective function J(x), where x is a vector consisting of the maximum compensation current value, current change rate threshold, and PID integral time constant in the preset parameter matrix. The objective function expression is: J(x) = w1*(Imax - Iref)^2 + w2*(dI / dt - dIref / dt)^2 + w3*(Ti - Tiref)^2 Among them, Imax, dI / dt, and Ti are the maximum compensation current value, the current change rate threshold, and the PID integral time constant, respectively; Iref, dIref / dt, and Tiref are the corresponding reference values; and w1, w2, and w3 are weight coefficients.

[0047] Constraints include: g1(x): |Rs - Rsref|<= ΔRs_max; g2(x): |L - Lref|<= ΔL_max; g3(x): |Ke - Keref|<= ΔKe_max; Among them, Rs, L, Ke are stator resistance, mutual inductance, and back electromotive force coefficient respectively, Rsref, Lref, Keref are corresponding reference values, ΔRs_max, ΔL_max, ΔKe_max are maximum allowable deviations.

[0048] The sequential quadratic programming algorithm is used to solve the optimization problem. Through iterative solution, the updated preset parameter matrix is ​​finally obtained, which includes the maximum compensation current value, current change rate threshold and PID integral time constant under different voltage fluctuation modes and scenario coding combinations.

[0049] Through the above technical solution, the present application realizes the accurate update of the preset parameter matrix under the constraints of dynamic grid fluctuations and multi-scenario noise. Since the sequential quadratic programming algorithm is used to solve the nonlinear constrained optimization problem, the problem of local optimal solution caused by linear weighting only by weight coefficients is avoided. At the same time, by constructing an objective function including the maximum compensation current value, the current change rate threshold and the PID integral time constant, and introducing the motor parameter drift and the drift weight coefficient as constraints, the coupling relationship between different control parameters is effectively balanced. This method improves the accuracy of the preset parameter matrix update, thereby achieving the global minimization of the motor control error, and enhancing the operating stability and quiet performance of the low-power distilled water machine in complex power grid environments and multi-noise scenarios.

[0050] In some of the above-mentioned schemes in the present application, a scheme is proposed to retrieve a preset parameter matrix through a combination of voltage fluctuation pattern and scene coding for dynamically adjusting control parameters. However, in a complex power grid environment, when there is a multi-dimensional parameter intersection in the combination of voltage fluctuation pattern and scene coding, the traditional matrix retrieval method may cause delays in control parameter acquisition due to the low efficiency of multi-dimensional data matching, thereby affecting the real-time performance of motor control. In this regard, the present application further proposes: determining the voltage level corresponding to the voltage fluctuation pattern and the noise level corresponding to the scene encoding; based on the voltage level and the noise level, matching is performed in an index table of a preset parameter matrix, and the index table stores a mapping relationship between the voltage level, the noise level and the parameter storage address; in response to the existence of a parameter storage address in the index table that matches both the voltage level and the noise level, reading the maximum compensation current value, the current change rate threshold and the PID integral time constant from the parameter storage address. Among them, the voltage level is divided according to the discretization interval of the voltage fluctuation amplitude. For example, the voltage fluctuation range of the power grid is divided into three levels: within ±5% is level 1, ±5%~10% is level 2, and more than ±10% is level 3. The noise level is determined based on the environmental noise threshold of the scene coding analysis. For example, the medical scene corresponds to level 1 low noise, the laboratory scene corresponds to level 2 medium noise, and the industrial scene corresponds to level 3 high noise. The index table adopts a two-dimensional hash structure, with the combination of voltage level and noise level as the key value, and directly maps to the parameter storage address. For example, when the voltage level is level 2 and the noise level is level 1, the hash key value can be encoded as "2-1", pointing to the corresponding physical storage address. The parameter storage address contains data blocks of the maximum compensation current value, the current change rate threshold, and the PID integral time constant. The size of each data block is fixed to 12 bytes, of which the first 4 bytes store the maximum compensation current value, the middle 4 bytes store the current change rate threshold, and the last 4 bytes store the PID integral time constant. Through the combination of discretization grading and hash mapping, the multi-dimensional parameter matching is converted into a two-dimensional key value query, avoiding traversing the entire parameter matrix, and the query time is shortened to milliseconds. Specifically, when the grid voltage fluctuation exceeds the threshold, the voltage fluctuation mode is classified into the corresponding voltage level. For example, when the voltage sag amplitude is detected to be 8%, it is determined to be the second-level voltage level. At the same time, the scene coding parsing unit reads the current environment RFID tag information, for example, the medical area tag corresponds to the first-level noise level. Based on the second-level voltage level and the first-level noise level, a hash key value "2-1" is generated in the index table, and the physical address of the combined parameter stored in the preset parameter matrix is ​​directly located. If there is a match in the index table, the maximum compensation current value 25A, the current change rate threshold 5A / ms and the PID integral time constant 0.2s are read from the address. Through the fast positioning of the hash index, the parameter retrieval time is reduced from 20ms of the traditional method to 2ms, meeting the real-time requirements of motor control. The read parameters act on the motor winding through the current gradient limiter and the PID controller, where the maximum compensation current value limits the current amplitude to reduce electromagnetic noise, the current change rate threshold suppresses current mutation, and the PID integral time constant adjusts the control response speed, thereby achieving stable operation of the motor under voltage fluctuation and multi-noise constraints.

[0051] As a preferred embodiment, the solution of the present application is specifically implemented as follows: In the magnetic levitation motor control system of the low-power distilled water machine, a voltage level classifier and a noise level classifier are set. The voltage level classifier divides the voltage fluctuation mode into 5 levels, corresponding to 0-5%, 5-10%, 10-15%, 15-20% and voltage fluctuation amplitudes above 20%. The noise level classifier divides the scene coding into 4 levels, corresponding to the noise limit requirements of 30dB, 40dB, 50dB and 60dB.

[0052] A 20×20 parameter index table is pre-established, with rows representing voltage levels and columns representing noise levels. Each table cell stores the parameter storage address under the corresponding conditions, pointing to a parameter set containing the maximum compensation current value, current change rate threshold, and PID integral time constant.

[0053] During operation, the voltage level classifier is used to determine the voltage level corresponding to the current voltage fluctuation mode. For example, if the voltage fluctuation amplitude is detected to be 12%, it is determined to be the voltage level 3. At the same time, the noise level classifier is used to determine the noise level corresponding to the current scene encoding. For example, if the scene is detected to be a medical environment, the corresponding noise limit is 40dB, and it is determined to be the noise level 2.

[0054] Next, locate the cell at row 3 and column 2 in the parameter index table and obtain the parameter storage address stored in the cell. The address points to a parameter set including a maximum compensation current value of 2.5A, a current change rate threshold of 0.5A / ms, and a PID integral time constant of 15ms.

[0055] Finally, the control system directly reads the three parameter values ​​from this address for subsequent motor control.

[0056] Through the above technical scheme, the present application realizes fast parameter matching under voltage fluctuation and noise restriction conditions. Compared with the traditional multi-dimensional matrix retrieval method, this scheme simplifies the complex multi-dimensional matching into an index table search and a parameter reading, which greatly improves the parameter acquisition speed. This efficient parameter matching mechanism ensures that when the grid voltage fluctuates rapidly, the control system can adjust the control parameters in time and effectively suppress the impact of voltage fluctuations on the operation of the motor. At the same time, by accurately matching the control parameters corresponding to the noise level, accurate noise control in different scenarios is achieved to meet diverse application needs. In addition, the index table structure of this scheme has good scalability, and new voltage levels and noise levels can be easily added to adapt to more complex application environments.

[0057] In some of the above-mentioned schemes of the present application, a technical solution of dynamically adjusting the current gradient limiting strategy is proposed to address the problem that in a low-power distilled water machine, under grid voltage fluctuations and multiple noise threshold scenarios, the motor winding current change rate exceeds the threshold, resulting in excessive electromagnetic noise, and the voltage fluctuation affects the stable operation of the motor. In this regard, the present application further proposes the following technical means: calculating the difference between the actual change rate of the motor winding current and the current change rate threshold; when the difference is greater than zero, starting the current gradient limiter, and determining the PWM duty cycle adjustment rate in response to the voltage fluctuation mode, the voltage sag mode corresponds to a fast adjustment rate, and the voltage slow change mode corresponds to a smooth adjustment rate; based on scene coding, determining the PWM switching frequency; adjusting the PWM duty cycle according to the PWM duty cycle adjustment rate, and driving the motor winding voltage at the PWM switching frequency, suppressing the actual change rate of the motor winding current, and making the actual change rate converge to the current change rate threshold. Among them, the starting condition of the current gradient limiter is realized by real-time difference calculation, and the difference calculation module can be integrated into the signal processing unit of the motor controller. The determination of the PWM duty cycle adjustment rate depends on the classification result of the voltage fluctuation mode. In the voltage sag mode, the adjustment rate with a millisecond response speed is adopted. For example, the duty cycle change rate is set to 2000 times per second, while it is adjusted to 500 times per second in the voltage slow fluctuation mode to match different dynamic requirements. The selection of PWM switching frequency is based on the noise level corresponding to the scene encoding. For example, the medical scene corresponds to high-frequency switching (such as above 20kHz), and the laboratory scene uses medium frequency (15kHz) to reduce audible noise through frequency switching. The application of PID integral time constant is further combined with the dynamic adjustment of compensation current. For example, the integral time is shortened to 10ms to accelerate the response when the voltage sags, and extended to 50ms to smooth the output when it fluctuates slowly. Specifically, after detecting that the actual rate of change of the current exceeds the threshold, the current gradient limiter is activated. According to the voltage fluctuation pattern recognition result, the corresponding duty cycle adjustment rate is selected: when the voltage drops suddenly, the duty cycle is adjusted quickly (for example, 2000 changes per second) to quickly reduce the winding voltage to suppress the current mutation; when the voltage fluctuates slowly, a gentle adjustment rate (for example, 500 changes per second) is adopted to avoid excessive adjustment and cause high-frequency noise. At the same time, the scene coding parsing unit obtains the noise limit requirements of the current scene. For example, the medical scene needs to increase the PWM switching frequency to above 20kHz so that the electromagnetic noise frequency band exceeds the sensitive range of the human ear. The adjusted duty cycle signal and the switching frequency act together on the motor drive circuit to limit the current change rate within the threshold. Furthermore, the PID controller corrects the output voltage in real time based on the maximum compensation current value and the integral time constant, for example, by shortening the integral time constant to speed up the compensation response and ensure the current stability under voltage fluctuations. Through the above steps, the response lag problem during voltage drops is avoided, and the electromagnetic noise in the gently fluctuating scene is reduced, achieving a balance between dynamic response and noise suppression. As a preferred embodiment, the solution of the present application is specifically implemented as follows: Calculate the difference between the actual rate of change of the motor winding current and the current rate of change threshold. The motor controller uses a high-precision current sampling circuit with a sampling frequency set to 20kHz. The actual current rate of change is calculated by the difference between two consecutive current sampling values. The current rate of change threshold is obtained by looking up a table, for example, it is set to 0.5A / ms in a medical scenario. When the difference is greater than zero, the current gradient limiter is activated. The current gradient limiter is implemented using FPGA. In response to the voltage fluctuation mode, the PWM duty cycle adjustment rate is determined. The voltage sag mode corresponds to a fast adjustment rate, which is set to 10% / ms; the voltage slow change mode corresponds to a gentle adjustment rate, which is set to 1% / ms. Based on the scene coding, the PWM switching frequency is determined. For example, the medical scene corresponds to 20kHz, the laboratory scene corresponds to 15kHz, and the office scene corresponds to 10kHz. The PWM duty cycle is adjusted according to the PWM duty cycle adjustment rate. The duty cycle adjustment is divided into 10 levels using a piecewise linear interpolation algorithm, and each level corresponds to an adjustment time of 0.1ms. The motor winding voltage is driven at the PWM switching frequency to suppress the actual change rate of the motor winding current so that the actual change rate converges to the current change rate threshold. Through the above technical scheme, the present application realizes the adaptive adjustment of the motor control strategy to the voltage fluctuation mode. In the case of a sudden voltage drop, the PWM duty cycle is adjusted quickly to effectively suppress the risk of motor instability; when the voltage changes slowly, a smooth adjustment strategy is adopted to avoid the introduction of additional electromagnetic noise. At the same time, the PWM switching frequency is dynamically adjusted through scene coding to achieve precise control of motor noise in different noise sensitivity scenarios. Therefore, while ensuring the stable operation of the motor, this solution meets the differentiated needs of noise control in different application scenarios, and improves the adaptability of low-power distilled water machines in complex power grid environments and multi-scenario applications.

[0058] In some of the above-mentioned schemes of the present application, a method of reducing the motor winding voltage by adjusting the PWM duty cycle to suppress the current change rate is proposed. However, in the specific implementation process, due to the differentiated requirements for the duty cycle adjustment rate of different voltage fluctuation modes (such as voltage sag and slow change), if a single adjustment rate is used, it will lead to response lag in the voltage sag scenario or excessive adjustment in the slow change scenario; at the same time, the noise levels corresponding to different scenario encodings have different sensitivities to the PWM switching frequency. If the switching frequency is not dynamically adjusted according to the scenario, it may cause the high-frequency switching noise to exceed the standard or the low-frequency switching to cause insufficient current suppression effect.

[0059] In this regard, the present application further proposes: when the difference is greater than zero, the current gradient limiter is started, and the PWM duty cycle adjustment rate is determined in response to the voltage fluctuation mode, the voltage drop mode corresponds to a fast adjustment rate, and the voltage slow change mode corresponds to a smooth adjustment rate; based on scene coding, the PWM switching frequency is determined; the PWM duty cycle is adjusted according to the PWM duty cycle adjustment rate, and the motor winding voltage is driven at the PWM switching frequency to suppress the actual change rate of the motor winding current, so that the actual change rate converges to the current change rate threshold.

[0060] Among them, the determination of the PWM duty cycle adjustment rate can be achieved by setting a multi-level rate threshold. For example, the duty cycle adjustment rate is set to 5%-10% per millisecond in the voltage sag mode, and adjusted to 0.5%-1% per millisecond in the voltage slow change mode; the determination of the PWM switching frequency is based on the preset noise level and frequency mapping table in the scene encoding. For example, the medical scene corresponds to a high frequency above 20kHz to avoid the sensitive frequency band of the human ear, and the industrial scene allows a low frequency below 15kHz to reduce switching losses. When the current gradient limiter is started, the microcontroller compares the actual current change rate with the threshold difference in real time. If the difference exceeds zero, the mode judgment logic is triggered, and the adjustment rate is switched in combination with the mode signal output by the voltage fluctuation detection module. The linkage control of the PWM duty cycle adjustment rate and the switching frequency is realized through a hardware timer. For example, a 1μs timer resolution is used in the fast adjustment mode, and a 10μs resolution is switched to reduce the calculation load in the smooth mode. The synergy between the duty cycle adjustment rate and the scene coding is reflected in the following: when the scene coding indicates a high noise-sensitive environment, even in the voltage sag mode, the switching frequency still prioritizes meeting the noise limit requirements, while compensating for the current suppression effect by increasing the duty cycle adjustment rate.

[0061] Specifically, after the current gradient limiter is started, the voltage fluctuation mode is monitored in real time. If the voltage change rate is detected to exceed 500V / s, it is determined to be a sudden drop mode. At this time, the duty cycle adjustment rate is set to a linear change rate of 8% per millisecond, and the PWM switching frequency is set to 22kHz according to the medical area tag information in the scene coding; when the voltage change rate is lower than 50V / s, it switches to a gentle mode, the duty cycle adjustment rate is reduced to 0.8% per millisecond, and the switching frequency is adjusted to 12kHz according to the industrial scene coding. The PWM duty cycle is achieved by adjusting the on-time ratio of the pulse width modulation signal. For example, in the sudden drop mode, the duty cycle is quickly reduced from 75% to 65% in only 1.25ms, while the same amplitude adjustment in the gentle mode takes 10ms. The adjusted PWM signal is applied to the motor winding through the drive circuit, so that the reduction amplitude of the winding voltage matches the adjustment rate, thereby controlling the current change rate within the threshold range. This dynamic adjustment mechanism reduces the current suppression response time to 1 / 5 of the traditional method when the voltage drops by eliminating the hysteresis effect of fixed parameter control, while reducing electromagnetic noise by more than 12dB in noise-sensitive scenarios.

[0062] As a preferred embodiment, the solution of the present application is specifically implemented as follows: In the magnetic levitation motor control system of the low-power distilled water machine, a voltage fluctuation detection module and a scene coding analysis module are set. The voltage fluctuation detection module uses a high-speed sampling circuit with a sampling frequency of 10kHz to detect the rate of change of the grid voltage. When the voltage change rate exceeds 100V / s, it is determined to be a voltage sag mode; when the voltage change rate is between 10-100V / s, it is determined to be a voltage slow change mode. The scene coding analysis module obtains the current scene code by reading the environmental RFID tag information. For example, the medical scene code is 01, the laboratory scene code is 02, and the office scene code is 03.

[0063] The control system determines the PWM duty cycle adjustment rate according to the voltage fluctuation mode. The PWM duty cycle adjustment rate corresponding to the voltage sag mode is 50% / ms, and the PWM duty cycle adjustment rate corresponding to the voltage slow change mode is 10% / ms. At the same time, the PWM switching frequency is determined based on the scene coding, with 20kHz for medical scenes, 15kHz for laboratory scenes, and 10kHz for office scenes.

[0064] The control system uses a digital signal processor (DSP) to achieve PWM duty cycle adjustment and switching frequency control. The DSP outputs a PWM waveform through a 12-bit DAC with a resolution of 0.024%. The PWM waveform is amplified by the drive circuit and drives the motor winding. The DSP monitors the motor winding current in real time with a sampling frequency of 100kHz. When it is detected that the actual current change rate exceeds the preset threshold, the DSP adjusts the PWM duty cycle according to the determined PWM duty cycle adjustment rate and drives the motor winding voltage at the determined PWM switching frequency.

[0065] For example, when a voltage sag occurs in a medical scenario, the DSP reduces the PWM duty cycle at a rate of 50% / ms and sets the PWM switching frequency to 20kHz. Rapid adjustment of the PWM duty cycle suppresses current mutations, and high-frequency switching effectively reduces electromagnetic noise. The DSP continuously monitors the current change rate until it converges to a preset threshold.

[0066] Through the above technical solutions, the present application realizes the adaptive adjustment of motor control parameters to grid voltage fluctuations and application scenarios. Through differentiated PWM duty cycle adjustment rates, voltage drops can be responded to quickly to avoid noise caused by current mutations; for slow voltage changes, a gentle adjustment rate is adopted to maintain the stability of motor operation. At the same time, the PWM switching frequency is dynamically adjusted according to the scene requirements to effectively control electromagnetic noise in different scenarios while ensuring the current suppression effect. This adaptive control mechanism breaks through the limitations of traditional fixed parameter control, and realizes the stable operation and dynamic noise control of the low-power distilled water machine magnetic levitation motor under grid voltage fluctuations and multiple noise threshold scenarios.

[0067] Reference Figure 2 The present application further proposes a magnetic levitation motor control device for controlling the operation of a magnetic levitation motor of a low-power distilled water machine in a scenario where the grid voltage fluctuation exceeds a threshold and multiple noise thresholds, including: The voltage detection unit 210 is used to detect the rate of change of the grid voltage and determine the voltage fluctuation mode based on a preset threshold value; The scene code parsing unit 220 is used to read the environmental RFID tag information and obtain the current scene code; A parameter retrieval unit 230 is used to retrieve a maximum compensation current value, a current change rate threshold and a PID integral time constant from a preset parameter matrix based on a voltage fluctuation mode and a scene code; The control execution unit 240 is used to limit the rate of change of the motor winding current according to the current change rate threshold value, and to perform motor control through the maximum compensation current value and the PID integral time constant.

[0068] By detecting the grid voltage fluctuation pattern and dynamically adjusting the control parameters in combination with scenario coding, current mutations can be suppressed and electromagnetic noise can be reduced. At the same time, the parameter matrix is ​​optimized according to load status and environmental factors to improve the operating stability of the motor under different working conditions.

[0069] In addition, in some preferred embodiments, a magnetic levitation motor control device proposed in the present application can perform any step of the above method.

[0070] The present application further proposes a magnetic levitation motor, to which the above control method is applied.

[0071] This application realizes the deep integration of the magnetic levitation motor control strategy and the motor body characteristics. This solution can dynamically adjust the control parameters according to the grid voltage fluctuations and operating scenarios, effectively solving the contradiction between multi-scenario silent control and stability under voltage fluctuations. By combining the control method with the physical structure of the magnetic levitation motor, a closed-loop matching mechanism between the control strategy and the motor body characteristics is constructed to ensure that the parameter update does not exceed the linear working area of ​​the magnetic levitation bearing, thereby realizing the coordinated optimization of control stability and multi-scenario silent constraints. This mechatronic design breaks through the limitations of the decoupling of traditional control strategies and motor body parameters, and improves the adaptability and performance of magnetic levitation motors in complex grid environments and multiple noise threshold scenarios.

[0072] The above description is only an embodiment of the present application and is not intended to limit the protection scope of the present application. For those skilled in the art, the present application may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A magnetic levitation motor control method, used for the operation control of the magnetic levitation motor of a low-power distilled water machine in a scenario where the grid voltage fluctuation exceeds a threshold and multiple noise thresholds, characterized in that: The following steps are involved: Detecting the rate of change of grid voltage and determining the voltage fluctuation mode based on a preset threshold; The scene code parsing unit reads the environmental RFID tag information to obtain the current scene code; Based on the voltage fluctuation pattern and scenario coding, the maximum compensation current value, current change rate threshold and PID integration time constant are retrieved from the preset parameter matrix; The motor winding current change rate is limited according to the current change rate threshold, and the motor control is performed by the maximum compensation current value and the PID integral time constant.

2. A magnetic levitation motor control method according to claim 1, characterized in that: The step of retrieving the maximum compensation current value, the current change rate threshold and the PID integral time constant from the preset parameter matrix based on the voltage fluctuation mode and the scene coding includes: Collecting motor operation data in real time, the motor operation data including motor winding temperature, motor speed and stator current; Calculating the motor parameter drift based on the motor winding temperature, motor speed and stator current, wherein the motor parameters include stator resistance, mutual inductance and back electromotive force coefficient; According to the motor parameter drift, the least square method is used to update the preset parameter matrix, wherein the preset parameter matrix includes the maximum compensation current value, the current change rate threshold value and the PID integral time constant under different voltage fluctuation modes and scene coding combinations; In response to changes in the voltage fluctuation mode and the scene coding, the updated preset parameter matrix is ​​queried to obtain the corresponding maximum compensation current value, current change rate threshold and PID integral time constant.

3. A magnetic levitation motor control method according to claim 2, characterized in that: The step of updating a preset parameter matrix using the least square method according to the motor parameter drift, wherein the preset parameter matrix includes the maximum compensation current value, the current change rate threshold and the PID integral time constant under different voltage fluctuation modes and scene coding combinations includes: Obtaining the current load status of the distilled water machine, wherein the load status includes water level and water temperature; Based on the water level and the water temperature, determining a motor parameter drift weight coefficient corresponding to the current load state, wherein the motor parameter drift weight coefficient includes a stator resistance weight coefficient, a mutual inductance weight coefficient, and a back electromotive force weight coefficient; According to the motor parameter drift amount and the motor parameter drift weight coefficient, the weighted least squares method is used to update the preset parameter matrix, which includes the maximum compensation current value, the current change rate threshold and the PID integral time constant under different voltage fluctuation modes and scene coding combinations.

4. A magnetic levitation motor control method according to claim 1, characterized in that: The step of detecting the grid voltage change rate and determining the voltage fluctuation mode based on a preset threshold comprises: Collect grid voltage data in real time and filter it using a sliding average filtering algorithm; Calculate the change rate of the filtered voltage data within a preset time window to obtain the voltage change rate; Perform spectrum analysis on the voltage change rate and extract the main frequency component of voltage fluctuation; When the voltage change rate exceeds a preset threshold and the main frequency component is lower than a preset frequency threshold, the voltage fluctuation mode is determined.

5. A magnetic levitation motor control method according to claim 3, characterized in that: The step of updating a preset parameter matrix using a weighted least square method according to the motor parameter drift amount and the motor parameter drift weight coefficient, wherein the preset parameter matrix includes a maximum compensation current value, a current change rate threshold value and a PID integral time constant under different voltage fluctuation modes and scene coding combinations includes: Constructing an objective function, wherein the objective function takes the maximum compensation current value, the current change rate threshold value and the PID integral time constant in the preset parameter matrix as independent variables, takes the motor parameter drift amount and the motor parameter drift weight coefficient as constraint conditions, and the objective function is used to minimize the motor control error; A sequential quadratic programming algorithm is used to solve the objective function and obtain an updated preset parameter matrix, which includes a maximum compensation current value, a current change rate threshold and a PID integral time constant under different voltage fluctuation modes and scene coding combinations.

6. A magnetic levitation motor control method according to claim 1, characterized in that: The step of retrieving the maximum compensation current value, the current change rate threshold and the PID integral time constant from the preset parameter matrix based on the voltage fluctuation mode and the scene coding includes: Determine the voltage level corresponding to the voltage fluctuation pattern and the noise level corresponding to the scene coding; Based on the voltage level and the noise level, matching is performed in an index table of a preset parameter matrix, wherein the index table stores a mapping relationship between the voltage level, the noise level and the parameter storage address; In response to the presence of a parameter storage address in the index table that matches both the voltage level and the noise level, a maximum compensation current value, a current change rate threshold, and a PID integral time constant are read from the parameter storage address.

7. A magnetic levitation motor control method according to claim 1, characterized in that: The steps of limiting the rate of change of the motor winding current according to the current change rate threshold and performing motor control through the maximum compensation current value and the PID integral time constant include: Calculate the difference between the actual rate of change of the motor winding current and the current rate of change threshold; When the difference is greater than zero, the current gradient limiter is started to reduce the motor winding voltage by adjusting the PWM duty cycle, thereby suppressing the actual change rate of the motor winding current and making the actual change rate converge to the current change rate threshold; Based on the maximum compensation current value, the target current value of the PID controller is calculated. According to the target current value and the PID integral time constant, the output voltage of the PID controller is adjusted to compensate for the influence of voltage fluctuation on the motor operation.

8. A magnetic levitation motor control method according to claim 7, characterized in that: The step of starting the current gradient limiter when the difference is greater than zero, reducing the motor winding voltage by adjusting the PWM duty cycle, suppressing the actual change rate of the motor winding current, and making the actual change rate converge to the current change rate threshold comprises: When the difference is greater than zero, the current gradient limiter is started, and the PWM duty cycle adjustment rate is determined in response to the voltage fluctuation mode, the voltage sag mode corresponds to a fast adjustment rate, and the voltage slow change mode corresponds to a gentle adjustment rate; Based on the scene code, determine the PWM switching frequency; The PWM duty cycle is adjusted according to the PWM duty cycle adjustment rate, and the motor winding voltage is driven at the PWM switching frequency to suppress the actual change rate of the motor winding current so that the actual change rate converges to the current change rate threshold.

9. A magnetic levitation motor control device, used for the operation control of the magnetic levitation motor of a low-power distilled water machine in a scenario where the grid voltage fluctuation exceeds a threshold and multiple noise thresholds, characterized in that: include: A voltage detection unit, used to detect the rate of change of the grid voltage and determine the voltage fluctuation mode based on a preset threshold; A scene code parsing unit is used to read the environmental RFID tag information and obtain the current scene code; A parameter retrieval unit, used to retrieve a maximum compensation current value, a current change rate threshold, and a PID integral time constant from a preset parameter matrix based on a voltage fluctuation mode and a scenario code; The control execution unit is used to limit the rate of change of the motor winding current according to the current change rate threshold value, and to perform motor control through the maximum compensation current value and the PID integral time constant.

10. A magnetic levitation motor, characterized in that: The magnetic levitation motor is applied with the method according to any one of claims 1 to 8.

Images