Control parameter adjusting method and system of magnetic suspension air feeder

Abstract

The invention discloses a control parameter adjusting method and system for a magnetic suspension air feeder, and relates to the technical field of high-precision control, and the method comprises the steps: constructing an information fusion model, inputting collected electromechanical thermal coupling features into the information fusion model, and generating electromechanical thermal feature vectors; using a vector decoupling operator to separate the electromechanical heat feature vector into electromechanical heat feature components, and combining the electromechanical heat feature components with real-time magnetic suspension blower operation data to generate an energy stable scalar value; constructing a distributed digital gene pool, mapping the energy stable scalar value into a discrete working condition label, matching the discrete working condition label with a gene segment in the distributed digital gene pool, and outputting an optimal gene segment; and analyzing the optimal gene segment to generate a control parameter set, and generating a multi-parameter change track through a smooth operation function group. According to the method, the distributed digital gene pool is constructed, so that the control parameters are adaptive to slow time-varying factors such as continuously changing operation conditions and equipment aging, and the robustness of the system is remarkably improved.

Description

Technical Field

[0001] This invention relates to the field of high-precision control technology, and in particular to a method and system for adjusting control parameters of a magnetic levitation blower. Background Technology

[0002] The control technology for magnetic levitation blowers has evolved from early single-loop PID regulation to a multi-physics collaborative optimization system. Sensorless control strategies achieve real-time rotor position estimation through back-EMF observation and sliding mode observers, significantly reducing reliance on displacement sensors. Adaptive control algorithms, combined with Lyapunov stability theory, can dynamically adjust gain parameters to suppress sudden load disturbances. In recent years, deep learning-based parameter tuning methods have utilized long short-time memory networks to mine historical operating data, achieving adaptive mapping of control parameters and improving the system's robustness under varying operating conditions.

[0003] In the current operation of magnetic levitation fans, the dynamic decoupling capability of the strong electromechanical-thermal coupling effect is insufficient, and the strong nonlinear interaction between electromagnetic field, mechanical vibration and thermal deformation is difficult to separate effectively, resulting in a mismatch between control parameters and actual physical state, which in turn leads to high-frequency oscillation and system instability risk. In addition, the working condition identification and parameter switching mechanism relies on the coarse-grained division of discrete working condition labels and static rule base matching, which cannot accurately respond to continuous time-varying working conditions, resulting in a continuous decay of control accuracy and a significant decrease in system robustness. Summary of the Invention

[0004] In view of the aforementioned existing problems, the present invention is proposed.

[0005] Therefore, this invention provides a method for adjusting the control parameters of a magnetic levitation blower to solve the problems of control parameter mismatch and operating condition switching delay.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution:

[0007] In a first aspect, the present invention provides a method for adjusting control parameters of a magnetic levitation blower, comprising: constructing an information fusion model; inputting collected electromechanical-thermal coupling features into the information fusion model to generate an electromechanical-thermal feature vector; using a vector decoupling operator to separate the electromechanical-thermal feature vector into electromechanical-thermal feature components, and combining them with real-time magnetic levitation blower operating data to generate a stable energy scalar value; constructing a distributed digital gene library; mapping the stable energy scalar value to discrete operating condition labels; matching the discrete operating condition labels with gene fragments in the distributed digital gene library to output the optimal gene fragment; and parsing the optimal gene fragment to generate control parameters. The system generates multi-parameter variation trajectories through a set of smoothing operation functions and performs high-frequency carrier modulation using a spatiotemporal crystal oscillation mechanism to generate proportional control signals and integral control signals. A time-sensitive network protocol is used to perform phase alignment between the proportional and integral control signals, achieving phase locking within a preset time window. A synchronization signal is generated and loaded into the execution path, which includes a main path and an auxiliary path. The synchronization signal is executed through the main path, and historical control parameter sets are called to the auxiliary path for simulation execution. Based on the execution results of the main path and the simulation results of the auxiliary path, the Lyapunov exponent difference value is calculated and the stability margin value is output.

[0008] When the stability margin value is within the preset safety range, a synchronization signal is output to the power drive unit; otherwise, the auxiliary path simulation result is output as the final execution instruction.

[0009] In a preferred embodiment of the control parameter adjustment method for the magnetic levitation fan described in this invention, the specific steps for generating a stable energy scalar value are as follows:

[0010] The collected mechanical, electrical, and thermodynamic features are coupled into electromechanical-thermal coupled features, which are then input into the constructed information fusion model to generate electromechanical-thermal feature vectors.

[0011] The electromechanical-thermal characteristic vector is separated into mechanical, electrical, and thermodynamic characteristic components by a vector decoupling operator, and then integrated into an electromechanical-thermal characteristic component.

[0012] By arbitrating the electromechanical and thermal characteristic components with the operating data of the magnetic levitation fan through multi-field coupling, a stable energy scalar value is generated.

[0013] As a preferred embodiment of the control parameter adjustment method for the magnetic levitation blower of the present invention, the specific steps for outputting the optimal gene fragment are as follows:

[0014] Each magnetic levitation blower device node generates a corresponding gene fragment based on historical operating data and attaches a device node identifier. It receives multiple gene fragments to form a distributed digital gene bank.

[0015] Discrete operating condition labels are generated using energy stability scalar values, and the feature similarity distance between the discrete operating condition labels and gene fragments is calculated. The gene fragment with the smallest feature similarity distance is selected as the optimal gene fragment.

[0016] As a preferred embodiment of the control parameter adjustment method for the magnetic levitation blower of the present invention, the specific steps for generating the proportional control signal and the integral control signal are as follows:

[0017] Input the optimal gene fragment into a preset parameter mapping table to extract the set of control parameters;

[0018] The control parameter set generates multi-parameter variation trajectories through linear interpolation functions and spline interpolation functions;

[0019] The spatiotemporal crystal oscillation mechanism is activated to generate a high-frequency carrier signal with a fixed frequency. The multi-parameter variation trajectory and the high-frequency carrier signal are then used to generate proportional control signals and integral control signals through analog modulation.

[0020] In a preferred embodiment of the control parameter adjustment method for the magnetic levitation blower described in this invention, the specific steps for generating a synchronization signal and loading it into the execution path are as follows:

[0021] A unified time reference is established for the nodes of the magnetic levitation blower equipment through time-sensitive networking protocol, and the zero-crossing time difference between the proportional control signal and the integral control signal is calculated as the phase difference.

[0022] Based on the phase difference, the signal transmission delay and phase response are adjusted to restore the orthogonality between the proportional control signal and the integral control signal;

[0023] The phase difference is continuously monitored within a fixed time window. When the phase difference does not exceed the phase difference locking range, the phase locking is determined to be complete, a synchronization signal is generated and loaded into the execution path.

[0024] As a preferred embodiment of the control parameter adjustment method for the magnetic levitation blower described in this invention, the specific steps for determining the output stability margin value are as follows:

[0025] The synchronization signal is transmitted to the dual-loop collaborative controller through the main path, and the rotor displacement and coil current are collected as the execution results of the main path.

[0026] The auxiliary path calls the historical control parameter set and loads it into the virtual dual-loop collaborative controller. Under the same discrete operating condition label, it outputs the simulated rotor displacement and simulated coil current as the simulation results of the auxiliary path.

[0027] Calculate the Lyapunov exponents for the execution results of the main path and the simulation results of the auxiliary path, and generate stability margin values ​​based on the calculation results.

[0028] In a preferred embodiment of the control parameter adjustment method for the magnetic levitation fan described in this invention, the output auxiliary path simulation result serves as the final execution instruction, and the specific steps are as follows.

[0029] The stability margin value is compared with the preset safety range. If the stability margin value is within the preset safety range, it is marked as a safe state; otherwise, it is marked as a risk state.

[0030] When the state is determined to be safe, the synchronization signal is transmitted to the power drive unit; when the state is determined to be risky, the auxiliary path simulation results are reconstructed into the final execution instruction and output to the power drive unit.

[0031] Secondly, this invention provides a control parameter adjustment system for a magnetic levitation blower, comprising: a feature extraction module, a gene matching module, a trajectory modulation module, a phase synchronization module, a channel verification module, and a decision execution module; the feature extraction module is used to construct an information fusion model, inputting the collected electromechanical-thermal coupling features into the information fusion model to generate an electromechanical-thermal feature vector; using a vector decoupling operator to separate the electromechanical-thermal feature vector into electromechanical-thermal feature components, and combining them with real-time magnetic levitation blower operating data to generate a stable energy scalar value; the gene matching module is used to construct a distributed digital gene library, mapping the stable energy scalar value to discrete operating condition labels, matching the discrete operating condition labels with gene fragments in the distributed digital gene library, and outputting the optimal gene fragment; the trajectory modulation module is used to parse the optimal gene fragment to generate a control parameter set. A multi-parameter variation trajectory is generated through a smoothing operation function set, and high-frequency carrier modulation is performed through a spatiotemporal crystal oscillation mechanism to generate proportional control signals and integral control signals. A phase synchronization module is used to perform phase alignment of the proportional control signals and integral control signals using a time-sensitive network protocol, achieving phase locking within a preset time window, generating a synchronization signal and loading it into the execution path, which includes a main path and an auxiliary path. A channel verification module is used to execute the synchronization signal through the main path, call historical control parameter sets to the auxiliary path for simulation execution, calculate the Lyapunov exponent difference value based on the main path execution result and the auxiliary path simulation result, and output a stability margin value. A decision execution module is used to output a synchronization signal to the power drive unit when the stability margin value is within a preset safety range; otherwise, it outputs the auxiliary path simulation result as the final execution instruction.

[0032] Thirdly, the present invention provides a computer device, including a memory and a processor, wherein the memory stores a computer program, wherein when the computer program is executed by the processor, it implements any step of the control parameter adjustment method for the magnetic levitation blower as described in the first aspect of the present invention.

[0033] Fourthly, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program is executed by a processor, it implements any step of the control parameter adjustment method for the magnetic levitation blower as described in the first aspect of the present invention.

[0034] The beneficial effects of this invention are as follows: by constructing a distributed digital gene library, the energy stability scalar value is dynamically mapped to discrete operating condition labels, enabling the control parameters to adaptively adapt to continuously changing operating conditions and slow time-varying factors such as equipment aging, thus significantly improving the robustness of the system; through the main and auxiliary path collaborative verification and risk decision-making mechanism, combined with the preset safety range, the real-time prediction of instability risk is realized, and under the risk state, the simulation results of the auxiliary path are reconstructed into continuous command output, completely eliminating transient overshoot and secondary oscillations during the operating condition switching process. Attached Figure Description

[0035] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0036] Figure 1 This is a flowchart illustrating the method for adjusting the control parameters of a magnetic levitation blower.

[0037] Figure 2 This is a schematic diagram of the control parameter adjustment system for a magnetic levitation blower.

[0038] Figure 3 A flowchart for generating control signals.

[0039] Figure 4 A flowchart for verifying the execution path and final instructions. Detailed Implementation

[0040] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0041] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0042] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.

[0043] Reference Figures 1-4 As one embodiment of the present invention, this embodiment provides a method for adjusting the control parameters of a magnetic levitation blower, including the following steps:

[0044] S1. Construct an information fusion model, input the collected electromechanical-thermal coupling features into the information fusion model to generate an electromechanical-thermal feature vector; use the vector decoupling operator to separate the electromechanical-thermal feature vector into electromechanical-thermal feature components, and combine them with real-time magnetic levitation fan operation data to generate a stable energy scalar value.

[0045] S1.1. Define a feature alignment layer based on the industrial real-time synchronization standard, an association calculation layer based on the device coupling matrix, and an output layer based on the law of energy conservation. Construct an information fusion model based on the feature alignment layer, the association calculation layer, and the output layer.

[0046] It should be noted that when defining the feature alignment layer, a clock alignment framework is constructed based on a precise time synchronization protocol. Standard unit conversion rules for mechanical vibration, current, and temperature are defined through a unified dimensional system, forming a spatiotemporally normalized constraint structure. The constraint structure is output to the definition stage of the associated computation layer. Based on the electromechanical-thermal interaction coefficient matrix of the equipment standard document, the bearing displacement, current harmonics, and temperature monitoring are divided into three major blocks according to the multiphysics coupling principle, generating a fusion computation rule framework. The fusion computation rule framework is then passed to the definition stage of the output layer. Based on the law of conservation of energy, physical constraint parameters such as bearing anisotropy weights, harmonic suppression coefficients, and thermal conduction compensation coefficients are solidified, forming a physically interpretable output structure. The constraint structure, the fusion computation rule framework, and the physically interpretable output structure are interconnected through standardized interfaces: the output interface of the feature alignment layer is directly connected to the input interface of the associated computation layer, and the output interface of the associated computation layer is directly connected to the input interface of the output layer, completing the construction of the information fusion model.

[0047] The equipment standard document is a collection of specifications formed by the magnetic levitation blower manufacturer during the R&D phase. It includes equipment node identifiers for the magnetic levitation blower, which adopt a three-segment coding structure (3-digit type code, 2-digit production line number, and 6-digit serial number). The first three digits of the type code directly define the equipment type, which is divided into high-speed type and energy-saving type. The mechanical assembly drawings indicate the spatial coordinates and tolerances of the bearings, stator and rotor. The electrical winding topology diagram specifies the number of coil turns and permanent magnet arrangement parameters. The thermodynamic simulation report specifies the heat conduction path of the heat sink. It also includes a set of core parameters determined by engineering calculations, such as the mechanical transmission efficiency coefficient table, electromagnetic coupling strength coefficient table, and heat conduction compensation coefficient table.

[0048] The multiphysics coupling principle describes the physical mechanism of the interaction between mechanical vibration field, electromagnetic field and temperature field during the operation of magnetic levitation fan. Specifically, the change in bearing displacement affects the stator winding current distribution through Lorentz force, the Joule heating effect of the winding current changes the temperature gradient of the heat sink, and the temperature rise of the heat sink reacts to the bearing gap through the thermal expansion of the material, forming a closed-loop energy transfer chain of displacement-current-temperature. The quantification of the multiphysics coupling principle depends on the simultaneous solution of Newton's second law of mechanical dynamics, Maxwell's equations of electromagnetic field and Fourier's law of thermal conduction in thermodynamics.

[0049] The information fusion model does not involve a traditional training process in industrial implementation. It is pre-fixed entirely based on the physical laws and design specifications of the magnetic levitation fan: the clock synchronization parameters of the feature alignment layer are directly determined by the microsecond-level alignment requirements of the precise time protocol, and the standard unit conversion rules are defined according to a unified dimensional system; the coupling matrix coefficients of the correlation calculation layer are derived from the mechanical transmission efficiency, electromagnetic coupling strength, and thermal conduction compensation coefficient tables in the equipment standard documents, and are quantified by the engineering calculation book and recorded in read-only memory; the energy conservation law of the output layer is calculated by the manufacturer based on the measured data of the bearing's triaxial kinetic energy distribution, the eddy current loss ratio at a specific harmonic frequency, and the verification results of the heat dissipation gradient efficiency, according to the first law of thermodynamics, combining the conservation of mechanical momentum and the electromagnetic field energy theorem, and is finally written into the permanent framework in one go through the production line calibration equipment. All parameters cannot be modified during the equipment's life cycle, and there are no gradient descent, backpropagation, or iterative optimization links throughout the process.

[0050] S1.2 Collect mechanical, electrical, and thermodynamic features and couple them into electromechanical-thermal coupled features. Input the electromechanical-thermal coupled features into the information fusion model and generate electromechanical-thermal feature vectors through time synchronization and cross-domain fusion.

[0051] It should be noted that when acquiring mechanical characteristics, the bearing's six-degree-of-freedom displacement signal is obtained through an eddy current displacement sensor, converted into a 0-10V analog voltage signal by a signal conditioning circuit, and quantized and stored as a mechanical vibration digital sequence at a 1kHz sampling rate. When acquiring electrical characteristics, a Hall current sensor is used to capture the three-phase winding current waveform, which is then processed by an anti-aliasing filter and quantized and stored as a current digital sequence at a 50kHz sampling rate. When acquiring thermodynamic characteristics, the temperature of 14 monitoring points is obtained through a PT100 platinum resistance temperature sensor, linearized by an RTD transmitter, and quantized and stored as a temperature digital sequence at a 10Hz sampling rate. The mechanical vibration digital sequence, current digital sequence, and temperature digital sequence are aligned according to the acquisition timestamp and then coupled into an electromechanical-thermal coupled feature.

[0052] The electromechanical-thermal coupling characteristics are input into the feature alignment layer of the information fusion model. Based on a precise time synchronization protocol, the time is unified to a microsecond-level time base. Mechanical vibration, current, and temperature are converted to their corresponding standard dimensions according to standard unit conversion rules, outputting three time-synchronized and dimensionally normalized signals. These three signals are then input into the correlation calculation layer, where a 32×32 solidified coupling matrix is ​​invoked to perform matrix multiplication, generating a 32-dimensional initial fusion vector. This 32-dimensional initial fusion vector is passed to the output layer, where physical constraint parameters are loaded according to the law of conservation of energy. These parameters are then injected into the 32-dimensional initial fusion vector through element-wise multiplication, outputting a 32-dimensional electromechanical-thermal feature vector, expressed as follows:

[0053] F = H⊙W·T(g(X));

[0054] Where F represents the 32-dimensional electromechanical-thermal feature vector; H represents the physical constraint parameter; W represents the solidification coupling matrix; T(·) represents the time synchronization function; g(·) represents the dimension normalization function; and X represents the electromechanical-thermal coupling feature.

[0055] S1.3. The electromechanical-thermal characteristic vector is separated into mechanical characteristic components, electrical characteristic components and thermodynamic characteristic components by a diagonal matrix type vector decoupling operator, and then integrated into electromechanical-thermal characteristic components.

[0056] It should be noted that the 32-dimensional electromechanical-thermal feature vector is input into the diagonal matrix vector decoupling operator, which is a 32×32-dimensional diagonal matrix. Through matrix multiplication, the mechanical feature components (1-6 elements forming a 6-dimensional displacement vector), electrical feature components (7-18 elements forming a 12-dimensional harmonic amplitude vector), and thermodynamic feature components (19-32 elements forming a 14-dimensional temperature vector) are separated. The separated mechanical, electrical, and thermodynamic feature components are then integrated into structured electromechanical-thermal feature components according to the equipment node identifier.

[0057] S1.4. The electromechanical and thermal characteristic components are combined with the operating data of the magnetic levitation fan through multi-field coupling arbitration to generate a stable energy scalar value.

[0058] It should be noted that the electromechanical-thermal characteristic components and the real-time operating data (output air pressure and power factor) of the magnetic levitation fan are input into the multi-field coupling arbitration mechanism. Based on the weight allocation rules, the basic weight values ​​of each physical quantity in the electromechanical-thermal characteristic components are extracted. The mechanical energy stability factor, electrical quality factor, and thermal balance factor are calculated using a weighted aggregation formula. The operating condition compensation coefficient is calculated by combining the power factor and output air pressure. The mechanical energy stability factor, electrical quality factor, thermal balance factor, and operating condition compensation coefficient are substituted into the energy conservation arbitration equation to generate an energy stability scalar value, expressed as follows:

[0059]

[0060] Where S represents the energy stability scalar value; M represents the mechanical energy stability factor; E represents the electrical energy quality factor; O represents the thermal balance factor; κ represents the operating condition compensation coefficient; ∈ represents the zero-prevention constant, with a fixed value of 10⁻⁶; and η represents the dimensionless normalization coefficient, calculated from the rated operating condition reference value.

[0061] It should also be noted that the weight allocation rule refers to allocating weights for the six-degree-of-freedom displacement of the bearing according to the proportion of kinetic energy in the three axes (xyz), defining the suppression coefficient for specific harmonic currents based on the proportion of eddy current loss, and assigning compensation factors to temperature monitoring points according to the requirements of heat conduction efficiency.

[0062] S2. Construct a distributed digital gene library, map energy stable scalar values ​​to discrete operating condition labels, match discrete operating condition labels with gene fragments in the distributed digital gene library, and output the optimal gene fragment.

[0063] S2.1 Each magnetic levitation blower device node generates a corresponding gene fragment based on historical operating data and adds a device node identifier. The central server receives multiple gene fragments to form a distributed digital gene bank.

[0064] It should be noted that each magnetic levitation blower device node extracts nearly 30 days of historical operating condition data (including bearing vibration spectrum, three-phase current waveform, and temperature distribution curve) from its local storage unit. Using a feature reduction algorithm, the historical operating condition data for each 10-minute period is compressed into a 128-dimensional feature vector as the basic data for gene fragments. Each gene fragment is then appended with a device node identifier and encapsulated as a structured gene fragment. Each magnetic levitation blower device node uploads the structured gene fragment to the central server via industrial Ethernet. After receiving the structured gene fragments uploaded by all magnetic levitation blower device nodes, the central server parses the device node identifier, distinguishes the device type based on the type code in the device node identifier, groups and stores the gene fragments according to the device type, and simultaneously establishes a device node identifier-gene fragment index table, forming a distributed digital gene library.

[0065] S2.2. Generate discrete operating condition labels by discretizing the energy stable scalar value. Calculate the feature similarity distance between the discrete operating condition label and the gene fragment in the distributed digital gene library, and select the gene fragment with the smallest feature similarity distance as the optimal gene fragment.

[0066] It should be noted that the operating condition threshold is defined based on the physical boundary and energy efficiency comprehensive analysis method: the mechanical load is gradually increased through magnetic levitation bearing overload test until the vibration sensor triggers a safety alarm, the energy stability scalar value at the alarm critical point is recorded, and a safety margin of three times the standard deviation is added to calculate the upper limit σ of the unstable operating condition; the wind turbine energy efficiency inflection point test is performed to identify the position of the energy stability scalar value corresponding to the optimal efficiency load rate, and the calibration error of the measuring instrument is subtracted to determine the lower limit of the stable operating condition. By performing Rockwell hardness tests on bearing alloy samples from the same batch, the standard deviation of the hardness values ​​of 30 bearing alloy samples was calculated and converted into scalar compensation values ​​according to the hardness-life decay correlation formula. The critical energy stability scalar value at which temperature rise leads to a 50% life decay in the fatigue life test of bearing alloy materials was recorded, and the scalar compensation value was superimposed. After calibration, the lower limit of the transition condition was obtained. The scalar value ranges of the upper limit of the unstable condition, the lower limit of the transition condition, and the lower limit of the stable condition were combined to form the condition level threshold. 0.0~σ represents the unstable condition, and the discrete condition is labeled 000. For transitional operating conditions, the discrete operating condition label is 001; For stable operating conditions, the discrete operating condition label is 010;

[0067] Input the energy stability scalar value into the discretized encoder to generate a 3-bit discrete operating condition label based on the operating condition level threshold; load the gene fragment corresponding to the target equipment type (e.g., high-speed equipment retrieval partition A) into the distributed digital gene library, and directly extract the 128-dimensional feature vector of each gene fragment; calculate the feature similarity distance between the discrete operating condition label and the 128-dimensional feature vector of each gene fragment, use the minimum value selector to filter the storage address of the gene fragment corresponding to the minimum feature similarity distance, and output the optimal gene fragment.

[0068] S3. Analyze the optimal gene fragment to generate a set of control parameters, generate a multi-parameter change trajectory through a smoothing operation function group, and perform high-frequency carrier modulation through a spatiotemporal crystal oscillation mechanism to generate proportional control signals and integral control signals;

[0069] S3.1 Input the optimal gene fragment into the parameter mapping table and extract the displacement control loop parameter set and the current control loop parameter set;

[0070] It should be noted that the process involves collecting the original waveforms of bearing displacement and instantaneous values ​​of stator winding current of the magnetic levitation fan under rated load and overload test conditions; performing time-frequency conversion on the original bearing displacement waveform to obtain the average displacement amplitude, forming the displacement frequency domain characteristic value; performing harmonic decomposition on the instantaneous value of stator winding current to extract the fundamental wave, forming the current harmonic characteristic value; combining the displacement frequency domain characteristic value and the current harmonic characteristic value with multi-physics coupling simulation to generate displacement control loop parameter sets and current control loop parameter sets: inputting the displacement frequency domain characteristic value into the mechanical vibration field simulation unit, calculating the rotor stiffness matrix and damping coefficient based on the bearing dynamics equation, and outputting the displacement control loop parameter set; inputting the current harmonic characteristic value into the electromagnetic field simulation unit, establishing control equations based on Maxwell's equations and the stator winding, discretizing the spatial electromagnetic field distribution using the finite element method, calculating the eddy current loss density and outputting the potential eddy current loss distribution, generating the current control loop parameter set; establishing feature vector index addresses, and solidifying the displacement control loop parameter set, current control loop parameter set, and feature vector index addresses into a parameter mapping table;

[0071] The optimal gene fragment is input into the parameter mapping table, and the feature vector index address is matched by the type code. The displacement control loop parameter set and the current control loop parameter set are extracted based on the feature vector index address.

[0072] S3.2 The displacement control loop parameter set and the current control loop parameter set generate multi-parameter variation trajectories through linear interpolation functions and spline interpolation functions;

[0073] It should be noted that the displacement control loop parameters are processed by a linear interpolation function, using a 1ms control period as the time base, and the displacement control loop parameters of adjacent control periods are linearly transitioned to generate a continuous trajectory of displacement parameters. The current control loop parameters are processed by a cubic spline interpolation function, and a harmonic order sequence is defined according to the electromagnetic characteristics of the motor. Using the harmonic order sequence as the node vector, B-spline control points are inserted between adjacent harmonic orders to generate a smooth trajectory of current parameters. The continuous trajectory of displacement parameters and the smooth trajectory of current parameters are aligned according to a precise time protocol and merged into a multi-parameter variation trajectory.

[0074] S3.3 Activate the spacetime crystal oscillation mechanism to generate a fixed-frequency high-frequency carrier signal, and generate proportional control signal and integral control signal by analog modulation method using the multi-parameter variation trajectory and the high-frequency carrier signal.

[0075] It should be noted that the spacetime crystal oscillation mechanism is activated, and a high-frequency carrier signal with a fixed frequency (center frequency 2MHz, stability ±1ppm) is generated through a phase-locked loop circuit. The product of the continuous trajectory of the displacement parameter in the multi-parameter variation trajectory and the high-frequency carrier signal is used to generate a proportional control signal (0-10V sine wave). The smooth trajectory of the current parameter is input into the integral modulation circuit to perform time-varying integral operation on the high-frequency carrier, generating an integral control signal (-5V to +5V triangular wave). The high-frequency components of the carrier are filtered out by a low-pass filter, and a smooth proportional control signal and integral control signal are output.

[0076] S4. The proportional control signal and the integral control signal are phase aligned using a time-sensitive network protocol, phase locking is achieved within a preset time window, a synchronization signal is generated and loaded into the execution path, the execution path including a main path and an auxiliary path;

[0077] S4.1 Establish a unified time reference for the nodes of the magnetic levitation fan equipment through time-sensitive networking protocol, and calculate the zero-crossing time difference between the proportional control signal and the integral control signal as the phase difference;

[0078] It should be noted that a time-sensitive network protocol is deployed in each magnetic levitation blower equipment node, and a nanosecond-level clock distribution mechanism is used to establish a unified time reference for all equipment nodes. The proportional control signal and the integral control signal are input to a high-speed voltage comparison circuit. When the proportional control signal crosses the amplitude midpoint threshold (5.0V), the current timestamp is recorded. When the integral control signal crosses the phase reference threshold (0.0V), the current timestamp is recorded. The phase difference is generated by calculating the absolute value of the time difference between the two timestamps.

[0079] It should also be noted that Time-Sensitive Networking Protocol (TSN) refers to a communication framework for achieving microsecond-level time synchronization in industrial control networks. Its core is to generate nanosecond-level time reference signals through a precision clock source, eliminate transmission path differences by using a point-to-point delay compensation mechanism, and reduce clock deviations of all network devices based on a master-slave clock synchronization algorithm.

[0080] The proportional control signal is a unipolar sine wave. When the proportional control signal rises from below 5V to above 5V, it indicates that the sine wave has entered the positive half-cycle from the negative half-cycle. The midpoint of the amplitude of the proportional control signal (midpoint of the 0-10V range, 5.0V) is the amplitude midpoint threshold. The integral control signal is a bipolar triangular wave. The true zero-crossing point is located at the voltage zero point. When the integral control signal crosses the voltage zero point from a negative value, it indicates that the triangular wave has entered the rising edge from the falling edge. The voltage zero point is the phase reference threshold.

[0081] S4.2. Based on the phase difference, dynamically adjust the signal transmission delay and phase response to restore the orthogonality between the proportional control signal and the integral control signal;

[0082] It should be noted that the phase difference is input to the proportional-integral (PI) controller, and the PPI controller output is divided into two paths: one path inputs the phase difference to a 12-bit digital-to-analog converter, converting it into a 0-3.3V analog voltage signal. This analog voltage signal drives a capacitor array consisting of 64 binary weighted capacitors. Each weighted capacitor is connected to the analog voltage signal transmission path via a CMOS switch. When the voltage increases, the number of switches increases, causing the total connected capacitance value to increase linearly. According to the transmission delay formula, this extends the propagation time of the proportional control signal in the transmission line. The adjusted proportional control signal is then buffered and isolated by two stages of emitter followers before being output, generating a delay-compensated proportional control signal.

[0083] Another path applies the correction voltage output from the proportional-integral controller to the cathodes of four parallel varactor diodes, controlling the total junction capacitance through reverse bias changes; the four varactor diodes are connected in series to the physical transmission path of the integral control signal, forming an LC phase-shifting structure with a fixed inductor (determined based on the resonant frequency inequality, with a value of 22μH), changing the phase of the integral control signal in real time according to the phase shift formula, and outputting an integral control signal with precise phase shift;

[0084] The delayed-compensated proportional control signal and the phase-precisely offset integral control signal are input into an analog multiplier to perform instantaneous multiplication, generating an AC signal containing sum and difference frequency components. This AC signal is then fed into a fourth-order RC low-pass filter, and the DC voltage is extracted through integration, expressed as follows:

[0085]

[0086] V1(t)=δ v1 +δv1 ·sin(2π·10 4 t);

[0087] V2(t)=δ v2 ·sawtooth(2π·10 4 t)

[0088] Where V0 represents the DC voltage; Q represents the signal period, with a value of 0.1 ms; t represents a certain moment; V1(t) represents the instantaneous voltage of the proportional control signal at a certain moment t; V2(t) represents the instantaneous voltage of the integral control signal at a certain moment t; δ v1 This represents the amplitude constant of the proportional control signal, with a value of 5; δ v2 The value represents the peak value of the integral control signal, and its value is 5; sawtooth(·) represents the asymmetric sawtooth wave function, which is used to precisely control the phase offset of the integral signal;

[0089] The phase difference Δθ is generated by calculating the DC voltage using the anticosine function, and the expression is:

[0090]

[0091] Where Δθ represents the phase difference;

[0092] When the phase difference Δθ is within the phase difference locking range (90°±0.1°, based on the hardware detection limit and stability boundary definition), the analog voltage signal and the correction voltage are locked, and the orthogonal relationship between the proportional control signal and the integral control signal is reconstructed.

[0093] S4.3 Continuously monitor the phase difference between the proportional control signal and the integral control signal within a fixed time window. When the phase difference does not exceed the phase difference locking range, determine that the phase locking is complete, generate a synchronization signal and load it into the execution path.

[0094] It should be noted that the phase difference Δθ is input to the window comparator and compared with the phase difference locking range. When Δθ is continuously within the phase difference locking range, a 200ms timer counter is started, triggering an accumulation count once every signal period (0.1ms). If Δθ does not exceed the limit within 2000 consecutive periods (corresponding to 200ms), the phase locking is completed. When the count value of the accumulation counter reaches 2000, an overflow interrupt is triggered, generating an interrupt signal. The interrupt signal drives the output of the D flip-flop to jump from low level to high level, generating a rising edge synchronization signal and loading it into the main path of the execution path.

[0095] S5. Execute the synchronization signal through the main path, call the historical control parameter set to the auxiliary path for simulation execution, calculate the Lyapunov exponent difference value and output the stability margin value based on the main path execution result and the auxiliary path simulation result.

[0096] S5.1. The synchronization signal is transmitted to the magnetic bearing position loop controller and the motor drive current loop controller through the main path, and the rotor displacement and coil current are collected as the main path execution results.

[0097] It should be noted that the synchronization signal is transmitted to the serial peripheral interface data line of the magnetic bearing position loop controller and the controller area network bus transceiver of the motor drive current loop controller; the rising edge of the synchronization signal triggers the magnetic bearing position loop controller to execute the position loop activation command, and at the same time sends a pulse width modulation enable command to the motor drive current loop controller; the rotor displacement is collected in real time by the magnetic bearing eddy current displacement sensor and converted into a digital displacement signal by the analog-to-digital converter; at the same time, the coil current is collected by the motor phase current shunt resistor and the signal conditioning circuit and converted into a digital current signal by the high-precision analog-to-digital converter; the rotor displacement and coil current are aligned by a time synchronization protocol.

[0098] S5.2 Call the historical displacement control loop parameter set and the historical current control loop parameter set to the auxiliary path, and load them into the virtual position loop controller and the virtual current loop controller to build a virtual execution environment;

[0099] It should be noted that the historical displacement control loop parameter set (including proportional coefficient, integral time and differential gain parameters) and the historical current control loop parameter set (including proportional coefficient, integral time and feedforward gain parameters) stored in the distributed digital gene bank are called and transmitted to the auxiliary path memory buffer through the high-speed serial bus; the historical displacement control loop parameter set is loaded into the virtual position loop controller, and the historical current control loop parameter set is loaded into the virtual current loop controller at the same time.

[0100] The virtual position loop controller instantiates three independent arithmetic units—a proportional multiplier, an integral accumulator, and a differential-differential unit—within a programmable logic device. It writes the proportional coefficients of the historical displacement control loop parameter set into the coefficient register of the proportional multiplier, the integral time into the clock divider register of the integral accumulator, and the differential gain parameter into the gain register of the differential-differential unit. The virtual current loop controller instantiates a proportional-integral (PI) arithmetic core and a feedforward compensator. It loads the proportional coefficients and integral time of the historical current control loop parameter set into the coefficient register group of the PI arithmetic core, and writes the feedforward gain parameter into the amplification coefficient register of the feedforward compensator. The proportional multiplier, integral accumulator, and differential-differential unit are interconnected via a parallel data bus to form a proportional-integral-differential arithmetic unit. The PI arithmetic core and the feedforward compensator are interconnected via an adder to form a proportional-integral-feedforward arithmetic unit. Finally, a cross-clock domain synchronous logic connection is configured between the PI arithmetic unit and the PI arithmetic unit to form a virtual execution environment.

[0101] It should also be noted that the relational database continuously collects historical operating data (including bearing vibration spectrum, three-phase current waveform, and temperature distribution curve) from the magnetic levitation blower equipment nodes. Data snapshots generated every 10 minutes are compressed into 128-dimensional feature vectors using a feature reduction algorithm. These vectors are then appended with equipment node identifiers (a three-segment encoding: 3-digit type code + 2-digit production line number + 6-digit serial number) to form structured gene fragments. These fragments are then uploaded to the central server via industrial Ethernet (EtherCAT, 100Mbps) and stored in groups according to equipment type (e.g., high-speed HS1, energy-saving ES2 partitions). Simultaneously, an index table is established using the equipment node identifier and timestamp as the primary key.

[0102] S5.3 Under the same discrete operating condition labeling conditions, the virtual position loop controller and the virtual current loop controller operate synchronously, and output simulated rotor displacement and simulated coil current as the auxiliary path simulation results;

[0103] It should be noted that the discrete operating condition label is input into the operating condition parser to retrieve the historical rotor displacement reference trajectory and historical coil current reference command corresponding to the discrete operating condition label; the virtual position loop controller receives the historical displacement reference trajectory as input, performs proportional-integral-differential operations based on the loaded historical displacement control loop parameter set, and generates the simulated rotor displacement; the virtual current loop controller synchronously receives the historical coil current reference command, performs proportional-integral-feedforward operations based on the historical current control loop parameter set, and outputs the simulated coil current; the operation process of the virtual position loop controller and the virtual current loop controller is triggered by a global clock synchronization signal, and the control cycle is updated once every microsecond; after the simulated rotor displacement and simulated coil current are encapsulated, a discrete operating condition label and a timestamp are attached, and they are stored in the auxiliary path result buffer to complete the output of the auxiliary path simulation result.

[0104] S5.4 Calculate the Lyapunov exponents of the main path execution results and the auxiliary path simulation results respectively, and generate stability margin values ​​based on the calculation results.

[0105] It should be noted that the rotor displacement and coil current acquired from the main path are aligned and matched to the same time base using timestamps, while the simulated rotor displacement and simulated coil current output from the auxiliary path are extracted. Phase space reconstruction is performed on the rotor displacement of the main path and the simulated rotor displacement of the auxiliary path, mapping the rotor displacement and simulated rotor displacement to a high-dimensional phase space respectively. The maximum Lyapunov exponent for the rotor displacement and simulated rotor displacement is calculated using a small data quantity method, and the displacement difference is obtained. Phase space reconstruction is performed on the coil current of the main path and the simulated coil current of the auxiliary path, mapping the coil current and simulated coil current to a high-dimensional phase space respectively. The maximum Lyapunov exponent for the coil current and simulated coil current is calculated, and the current difference is obtained. The displacement difference and current difference are used to generate displacement stability margin and current stability margin values ​​respectively through negative operations. The arithmetic mean of the displacement stability margin and current stability margin values ​​is calculated to generate the stability margin value. The expression for calculating the maximum Lyapunov exponent is as follows:

[0106]

[0107] Where λ represents the maximum Lyapunov exponent; t0 represents the initial time of phase space reconstruction; t1 represents the termination time of phase space reconstruction; N represents the total number of trajectory points in the phase space; k represents the index of the trajectory point in the phase space, with a value of 1-N; t k L(t) represents the time t corresponding to the k-th trajectory point; k L(t) represents the trajectory distance at time t; k-1 ) represents the trajectory distance at time t corresponding to the (k-1)th trajectory point.

[0108] S6. When the stability margin value is within the preset safety range, output a synchronization signal to the power drive unit; otherwise, output the auxiliary path simulation result as the final execution instruction.

[0109] S6.1 Compare the stability margin value with the preset safety range. If the stability margin value is within the preset safety range, mark it as a safe state; otherwise, mark it as a risk state.

[0110] It should be noted that, according to the Class A phase margin requirements in the mandatory stability standard for the magnetic bearing industry, the lower limit threshold of the safe range is defined as 0.5s. -1 The upper limit threshold of the safe range is determined by the maximum permissible overdamping rate of the magnetic bearing device and the high-frequency noise suppression requirements, which is 10s. -1 ;

[0111] When the stability margin is ≥0.5s -1 And ≤10s -1 If the stability margin value is less than 0.5s, then a safety status code "1" (safe state) is appended to the stability margin value in the status register; if the margin value is less than 0.5s... -1 or >10s-1 If so, a risk status code "0" (risk status) is appended to the stability margin value in the status register.

[0112] S6.2 When the state is determined to be safe, the synchronization signal is transmitted to the power drive unit; when the state is determined to be risky, the auxiliary path simulation result is reconstructed into the final execution instruction and output to the power drive unit.

[0113] It should be noted that when the status code in the status register is "1" (safe state), the synchronization signal is read and transmitted to the instruction register of the power drive unit via the SPI interface, triggering the power amplifier to unlock its output. When the status code in the status register is "0" (risk state), the simulated rotor displacement and simulated coil current are extracted from the auxiliary path result buffer, reconstructed into a continuous time series by cubic spline interpolation, input into the proportional-integral feedforward arithmetic unit of the virtual current loop controller to generate virtual control quantities, and then converted into three-phase drive commands through inverse Clarke transform, which are sent to the instruction register of the power drive unit via the CAN bus. Finally, the power drive unit executes the corresponding operation according to the type of received instruction to complete the control under the safe or risk state.

[0114] This embodiment also provides a control parameter adjustment system for a magnetic levitation fan, including: a feature extraction module, a gene matching module, a trajectory modulation module, a phase synchronization module, a channel verification module, and a decision execution module;

[0115] The feature extraction module is used to construct an information fusion model. It inputs the collected electromechanical-thermal coupling features into the information fusion model to generate an electromechanical-thermal feature vector. The vector decoupling operator is used to separate the electromechanical-thermal feature vector into electromechanical-thermal feature components, and combined with the real-time magnetic levitation fan operation data, an energy stability scalar value is generated.

[0116] The gene matching module is used to construct a distributed digital gene library, map energy stable scalar values ​​to discrete operating condition labels, match discrete operating condition labels with gene fragments in the distributed digital gene library, and output the optimal gene fragment.

[0117] The trajectory modulation module is used to analyze the optimal gene fragment to generate a set of control parameters, generate a multi-parameter change trajectory through a smoothing operation function group, and perform high-frequency carrier modulation through a spatiotemporal crystal oscillation mechanism to generate proportional control signals and integral control signals.

[0118] The phase synchronization module is used to perform phase alignment of the proportional control signal and the integral control signal using a time-sensitive network protocol, achieve phase locking within a preset time window, generate a synchronization signal and load it into the execution path, which includes a main path and an auxiliary path.

[0119] The channel verification module is used to execute the synchronization signal through the main path, call the historical control parameter set to the auxiliary path for simulation execution, calculate the Lyapunov exponent difference value and output the stability margin value based on the main path execution result and the auxiliary path simulation result;

[0120] The decision execution module is used to output a synchronization signal to the power drive unit when the stability margin value is within the preset safety range; otherwise, it outputs the auxiliary path simulation result as the final execution instruction.

[0121] This embodiment also provides a computer device applicable to the control parameter adjustment method for a magnetic levitation blower, comprising: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement the control parameter adjustment method for a magnetic levitation blower as proposed in the above embodiment.

[0122] The computer device can be a terminal, comprising a processor, memory, communication interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, carrier networks, NFC (Near Field Communication), or other technologies. The display screen can be an LCD screen or an e-ink screen. The input devices can be a touch layer covering the display screen, buttons, a trackball, or a touchpad on the computer device's casing, or an external keyboard, touchpad, or mouse.

[0123] This embodiment also provides a storage medium storing a computer program, which, when executed by a processor, implements the control parameter adjustment method for the magnetic levitation blower as proposed in the above embodiments. The storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read Only Memory (EPROM), Programmable Red-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk.

[0124] In summary, this invention significantly improves system robustness by: constructing a distributed digital gene library to dynamically map stable energy scalar values ​​to discrete operating condition labels, enabling control parameters to adaptively adapt to continuously changing operating conditions and slow time-varying factors such as equipment aging; and by using a primary and secondary path collaborative verification and risk decision-making mechanism, combined with a preset safety range, to achieve real-time prediction of instability risks, and by reconstructing the secondary path simulation results into continuous command outputs under risk conditions, thus completely eliminating transient overshoot and secondary oscillations during operating condition switching.

[0125] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for adjusting control parameters of a magnetic levitation blower, characterized in that: include, An information fusion model is constructed, and the collected electromechanical-thermal coupling features are input into the information fusion model to generate an electromechanical-thermal feature vector. The electromechanical-thermal feature vector is separated into electromechanical-thermal feature components using a vector decoupling operator, and combined with real-time magnetic levitation fan operation data to generate a stable energy scalar value. Construct a distributed digital gene bank, map energy stability scalar values ​​to discrete operating condition labels, match discrete operating condition labels with gene fragments in the distributed digital gene bank, and output the optimal gene fragment. The optimal gene fragment is analyzed to generate a set of control parameters. A multi-parameter change trajectory is generated through a set of smoothing operation functions. High-frequency carrier modulation is performed through a spatiotemporal crystal oscillation mechanism to generate proportional control signals and integral control signals. A time-sensitive network protocol is used to perform phase alignment between the proportional control signal and the integral control signal, phase locking is achieved within a preset time window, a synchronization signal is generated and loaded into the execution path, which includes a main path and an auxiliary path. The synchronization signal is executed through the main path, and the historical control parameter set is called to the auxiliary path for simulation execution. Based on the execution results of the main path and the simulation results of the auxiliary path, the Lyapunov exponent difference value is calculated and the stability margin value is output. When the stability margin value is within the preset safety range, a synchronization signal is output to the power drive unit; otherwise, the auxiliary path simulation result is output as the final execution instruction.

2. The control parameter adjustment method for the magnetic levitation fan as described in claim 1, characterized in that: The specific steps for generating a stable energy scalar value are as follows. The collected mechanical, electrical, and thermodynamic features are coupled into electromechanical-thermal coupled features, which are then input into the constructed information fusion model to generate electromechanical-thermal feature vectors. The electromechanical-thermal characteristic vector is separated into mechanical, electrical, and thermodynamic characteristic components by a vector decoupling operator, and then integrated into an electromechanical-thermal characteristic component. By arbitrating the electromechanical and thermal characteristic components with the operating data of the magnetic levitation fan through multi-field coupling, a stable energy scalar value is generated.

3. The method for adjusting the control parameters of a magnetic levitation blower as described in claim 2, characterized in that: The specific steps for outputting the optimal gene fragment are as follows: Each magnetic levitation blower device node generates a corresponding gene fragment based on historical operating data and attaches a device node identifier. It receives multiple gene fragments to form a distributed digital gene bank. Discrete operating condition labels are generated using energy stability scalar values, and the feature similarity distance between the discrete operating condition labels and gene fragments is calculated. The gene fragment with the smallest feature similarity distance is selected as the optimal gene fragment.

4. The control parameter adjustment method for the magnetic levitation fan as described in claim 3, characterized in that: The specific steps for generating the proportional control signal and the integral control signal are as follows. Input the optimal gene fragment into a preset parameter mapping table to extract the set of control parameters; The control parameter set generates multi-parameter variation trajectories through linear interpolation functions and spline interpolation functions; The spatiotemporal crystal oscillation mechanism is activated to generate a high-frequency carrier signal with a fixed frequency. The multi-parameter variation trajectory and the high-frequency carrier signal are then used to generate proportional control signals and integral control signals through analog modulation.

5. The control parameter adjustment method for the magnetic levitation blower as described in claim 4, characterized in that: The specific steps for generating the synchronization signal and loading it into the execution path are as follows. A unified time reference is established for the nodes of the magnetic levitation blower equipment through time-sensitive networking protocol, and the zero-crossing time difference between the proportional control signal and the integral control signal is calculated as the phase difference. Based on the phase difference, the signal transmission delay and phase response are adjusted to restore the orthogonality between the proportional control signal and the integral control signal; The phase difference is continuously monitored within a fixed time window. When the phase difference does not exceed the phase difference locking range, the phase locking is determined to be complete, a synchronization signal is generated and loaded into the execution path.

6. The method for adjusting control parameters of a magnetic levitation blower as described in claim 5, characterized in that: The specific steps for outputting the stability margin value are as follows. The synchronization signal is transmitted to the dual-loop collaborative controller through the main path, and the rotor displacement and coil current are collected as the execution results of the main path. The auxiliary path calls the historical control parameter set and loads it into the virtual dual-loop collaborative controller. Under the same discrete operating condition label, it outputs the simulated rotor displacement and simulated coil current as the simulation results of the auxiliary path. Calculate the Lyapunov exponents for the execution results of the main path and the simulation results of the auxiliary path, and generate stability margin values ​​based on the calculation results.

7. The method for adjusting control parameters of a magnetic levitation blower as described in claim 6, characterized in that: The output of the secondary path simulation result is used as the final execution instruction. The specific steps are as follows. The stability margin value is compared with the preset safety range. If the stability margin value is within the preset safety range, it is marked as a safe state; otherwise, it is marked as a risk state. When the state is determined to be safe, the synchronization signal is transmitted to the power drive unit; When a risk state is determined, the simulation results of the auxiliary path are reconstructed into the final execution instruction and output to the power drive unit.

8. A control parameter adjustment system for a magnetic levitation blower, based on the control parameter adjustment method for a magnetic levitation blower according to any one of claims 1 to 7, characterized in that: It includes a feature extraction module, a gene matching module, a trajectory modulation module, a phase synchronization module, a channel verification module, and a decision execution module; The feature extraction module is used to construct an information fusion model. It inputs the collected electromechanical-thermal coupling features into the information fusion model to generate an electromechanical-thermal feature vector. The vector decoupling operator is used to separate the electromechanical-thermal feature vector into electromechanical-thermal feature components, and combined with the real-time magnetic levitation fan operation data, an energy stability scalar value is generated. The gene matching module is used to construct a distributed digital gene library, map energy stable scalar values ​​to discrete operating condition labels, match discrete operating condition labels with gene fragments in the distributed digital gene library, and output the optimal gene fragment. The trajectory modulation module is used to analyze the optimal gene fragment to generate a set of control parameters, generate a multi-parameter change trajectory through a smoothing operation function group, and perform high-frequency carrier modulation through a spatiotemporal crystal oscillation mechanism to generate proportional control signals and integral control signals. The phase synchronization module is used to perform phase alignment of the proportional control signal and the integral control signal using a time-sensitive network protocol, achieve phase locking within a preset time window, generate a synchronization signal and load it into the execution path, which includes a main path and an auxiliary path. The channel verification module is used to execute the synchronization signal through the main path, call the historical control parameter set to the auxiliary path for simulation execution, calculate the Lyapunov exponent difference value and output the stability margin value based on the main path execution result and the auxiliary path simulation result; The decision execution module is used to output a synchronization signal to the power drive unit when the stability margin value is within the preset safety range; otherwise, it outputs the auxiliary path simulation result as the final execution instruction.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the processor executes the computer program, it implements the steps of the control parameter adjustment method for the magnetic levitation blower according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by the processor, it implements the steps of the control parameter adjustment method for the magnetic levitation blower according to any one of claims 1 to 7.

Images