PLC-based control method and system for high-pressure-ratio magnetic suspension centrifugal heat pump

Abstract

The application discloses a high-pressure-ratio magnetic suspension centrifugal heat pump control method and system based on PLC. The method comprises the following steps: collecting running data and PLC configuration information to generate a dynamic signal matrix through adaptive sampling rate adjustment; identifying a key pressure ratio control loop, extracting a vibration envelope and a surge precursor frequency to generate early warning characteristic parameters; reversely adjusting a magnetic bearing stiffness damping ratio to determine an anti-surge protection boundary; monitoring fin temperature difference detection information, generating a partition defrosting map through asymmetric frosting rate analysis, executing minimum energy consumption analysis to determine a reversing valve switching timing; performing running feature and oil-gas migration coupling analysis to generate oil-gas separation optimization parameters, adjusting a magnetic suspension gap to extract a zero-power-consumption suspension interval, and constructing a hierarchical energy-saving control configuration; generating a preventive maintenance time window through a magnetic bearing current ripple, establishing a fault-tolerant degradation operation mode and an adaptive control strategy, and finally generating a heat pump control instruction to realize intelligent control of the high-pressure-ratio magnetic suspension centrifugal heat pump.

Description

TECHNICAL FIELD

[0001] The application relates to the technical field of heat pump control, in particular to a high-pressure-ratio magnetic suspension centrifugal heat pump control method and system based on PLC. BACKGROUND

[0002] As a new generation of high-efficiency energy-saving equipment, the high-pressure-ratio magnetic suspension centrifugal heat pump can realize stable refrigeration and heating performance under extreme working conditions, and is widely used in harsh environments such as data center cooling, industrial waste heat recovery and polar heating. However, due to its complex and variable operating conditions, it involves magnetic suspension bearing control, compressor anti-surge protection, oil-gas separation optimization and other technical links, and the traditional control method is difficult to realize the coordinated optimization of each subsystem.

[0003] The existing heat pump control technology mainly relies on single parameter independent control strategy, lacks deep mining and multi-parameter coupling analysis capability of operation data. The traditional method cannot effectively handle the coupling relationship between the dynamic response characteristics of the magnetic suspension bearing and the working condition changes of the compressor, and it is also difficult to realize early identification of surge precursor and intelligent scheduling of preventive maintenance. Especially in the process of complex working condition switching, the nonlinear coupling effect between the system components leads to a significant decline in control performance, affecting the long-term stable operation and energy-saving effect of the equipment. SUMMARY

[0004] The application discloses a high-pressure-ratio magnetic suspension centrifugal heat pump control method and system based on PLC, which realizes multi-dimensional feature extraction and coupling analysis of operation data by constructing a dynamic signal matrix and an adaptive weight domain, combines a preventive maintenance time window and a hierarchical energy-saving control configuration, establishes a fault-tolerant degraded operation mode and an adaptive control strategy, and finally generates a comprehensive control instruction through multi-parameter fusion optimization, thereby providing an all-round intelligent control solution for the high-pressure-ratio magnetic suspension centrifugal heat pump.

[0005] The application discloses a high-pressure-ratio magnetic suspension centrifugal heat pump control method based on PLC, which comprises the following steps:

[0006] Collecting operation data and PLC configuration information of the high-pressure-ratio magnetic suspension centrifugal heat pump, and generating a dynamic signal matrix by adaptively adjusting the sampling rate of the operation data;

[0007] Identifying a key pressure ratio control loop through the dynamic signal matrix, extracting a vibration envelope and a surge precursor frequency from the key pressure ratio control loop to generate a warning feature parameter, and determining an anti-surge protection boundary by reversely adjusting a magnetic bearing stiffness damping ratio using the warning feature parameter;

[0008] The fin temperature difference detection information of the anti-surge protection boundary is monitored, the fin temperature difference detection information is subjected to asymmetric frosting rate analysis through the PLC configuration information to generate a partition defrosting map, and minimum energy consumption analysis is performed based on the partition defrosting map to determine the reversing valve switching timing;

[0009] The reversing valve switching timing is used for operation characteristic and oil-gas migration coupling analysis with the dynamic signal matrix to generate oil-gas separation optimization parameters, the magnetic suspension gap is adjusted according to the oil-gas separation optimization parameters to extract a zero-power-consumption suspension interval, and a hierarchical energy-saving control configuration is constructed according to the zero-power-consumption suspension interval.

[0010] The operation data are subjected to magnetic bearing current ripple analysis to predict the development trend of compressor eccentricity to generate a preventive maintenance time window, the preventive maintenance time window and the hierarchical energy-saving control configuration are matched to determine a self-adaptive control strategy, and the self-adaptive control strategy is used for multi-parameter fusion optimization to generate a heat pump control instruction.

[0011] The second aspect of the application provides a PLC-based high-pressure-ratio magnetic suspension centrifugal heat pump control system, comprising:

[0012] A data acquisition module is configured to acquire operation data and PLC configuration information of a high-pressure-ratio magnetic suspension centrifugal heat pump, and to generate a dynamic signal matrix by adaptively adjusting the sampling rate of the operation data.

[0013] An anti-surge control module is configured to identify a key pressure ratio control loop through the dynamic signal matrix, extract a vibration envelope and a surge precursor frequency from the key pressure ratio control loop to generate a warning characteristic parameter, and determine an anti-surge protection boundary by reversely adjusting a magnetic bearing stiffness-damping ratio using the warning characteristic parameter.

[0014] A timing switching module is configured to monitor fin temperature difference detection information of the anti-surge protection boundary, generate a partition defrosting map by performing asymmetric frosting rate analysis on the fin temperature difference detection information through the PLC configuration information, and determine the reversing valve switching timing based on minimum energy consumption analysis of the partition defrosting map.

[0015] An energy-saving optimization module is configured to use the reversing valve switching timing and the dynamic signal matrix to perform operation characteristic and oil-gas migration coupling analysis to generate oil-gas separation optimization parameters, adjust the magnetic suspension gap according to the oil-gas separation optimization parameters to extract a zero-power-consumption suspension interval, and construct a hierarchical energy-saving control configuration according to the zero-power-consumption suspension interval.

[0016] A prediction control module is configured to perform magnetic bearing current ripple analysis on the operation data to predict compressor eccentricity development trend, generate a preventive maintenance time window, match the preventive maintenance time window with a fault-tolerant degraded operation mode of the hierarchical energy-saving control configuration to determine an adaptive control strategy, and generate a heat pump control instruction based on the adaptive control strategy.

[0017] The beneficial effects of the present application are embodied in the following aspects: first, by using adaptive sampling rate adjustment and excitation-displacement coupling analysis technology, the control accuracy of the magnetic suspension bearing is effectively improved, the accurate control of variable density sampling points can capture the nonlinear characteristics of the bearing, combined with vibration envelope analysis and non-integer frequency feature recognition, these signal features are used for early warning of compressor surge, compared with the traditional passive protection mode, the active anti-surge technology expands the safe operation range of the compressor under high pressure ratio working condition, and improves the operation stability and reliability of the equipment under extreme working condition. Second, by using the zoning defrosting map and minimum energy consumption analysis technology, the defrosting efficiency of the heat exchanger is comprehensively optimized, through peak-shaving defrosting and intelligent timing control, the defrosting energy consumption is effectively reduced and the defrosting time is shortened, while the stability of the heat exchange efficiency is maintained, combined with the analysis and utilization of the lubricating oil redistribution phenomenon during reversing, through the extraction of zero-power suspension interval and hierarchical energy-saving control configuration, the power consumption of the magnetic suspension bearing is effectively reduced, and the operation efficiency of the whole machine is improved, and good energy-saving effect is realized under the premise of ensuring performance. Finally, by using current ripple analysis and preventive maintenance time window technology, the intelligent prediction ability of the equipment health state is established, the deterioration signals such as current ripple and mechanical eccentricity are fully utilized for fault trend analysis, the accuracy of fault prediction is improved, the implementation of preventive maintenance reduces the unplanned downtime of the equipment, and the maintenance cost is controlled, and the fault-tolerant degraded operation mode ensures that the equipment can still maintain basic functions during maintenance, through the adaptive control strategy of multi-parameter fusion optimization, the coordinated control between sub-components is realized, and the overall response speed is improved.

[0018] It should be understood that the above general description and the following detailed description are only exemplary and explanatory, and cannot limit the present application. BRIEF DESCRIPTION OF DRAWINGS

[0019] The drawings herein show specific examples of the technical solutions described in the present application, and constitute part of the specification together with the specific embodiments, for explaining the technical solutions, principles and effects of the present application.

[0020] Unless specifically stated, the same reference signs in different drawings represent the same or similar technical features, and different reference signs may also be used to represent the same or similar technical features.

[0021] Figure 1It is a flow chart of the control method of the high-pressure-ratio magnetic suspension centrifugal heat pump based on PLC.

[0022] Figure 2 It is a structure block diagram of the control system of the high-pressure-ratio magnetic suspension centrifugal heat pump based on PLC. DETAILED DESCRIPTION

[0023] In the following description, for purposes of explanation and not limitation, specific details are set forth such as particular architectures, technologies, techniques, etc. in order to provide a thorough understanding of the embodiments of the present application. However, it will be apparent to those skilled in the art that the present application can be practiced in other embodiments that depart from these specific details. In other instances, detailed descriptions of well-known systems, devices, circuits, and methods are omitted so as not to obscure the description of the present application with unnecessary detail.

[0024] It is to be understood that the terminology "including", "comprising", "consisting" and "consisting essentially of" used in the specification and the appended claims, indicates the presence of the stated features, integers, steps, operations, elements, and / or components but does not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or groups thereof.

[0025] Reference throughout this specification to "one embodiment" or "an embodiment" means that a particular feature, structure, or characteristic described in connection with the embodiment is included in at least one embodiment of the application. Thus, the appearances of the phrases "in one embodiment" or "in an embodiment" or "in other embodiments" or "in some embodiments" in various places throughout this specification are not necessarily referring to the same embodiment, unless otherwise specified. The terms "including", "comprising", "consisting essentially of" and "consisting of" are used interchangeably, unless otherwise specified.

[0026] The technical solutions of the embodiments of the present application are introduced as follows.

[0027] As shown in Figure 1 The embodiment of the present application provides a control method of a high-pressure-ratio magnetic suspension centrifugal heat pump based on PLC, which comprises the following steps S110-S150:

[0028] In step S110, the running data and the PLC configuration information of the high-pressure-ratio magnetic suspension centrifugal heat pump are collected, and the running data is adaptively sampled to generate a dynamic signal matrix.

[0029] Specifically, the system collects operational data and PLC configuration information for the high-pressure magnetic levitation centrifugal heat pump. A distributed data acquisition system synchronously collects operational data from the magnetic levitation bearing system, centrifugal compressor, and control system, while simultaneously reading PLC configuration parameters and control logic information. The operational data collection covers the excitation current, radial displacement, axial displacement, and vibration signals of the magnetic levitation bearing, as well as the speed, pressure, temperature, and flow parameters of the centrifugal compressor. Excitation current data is acquired using a high-precision current sensor with a sampling frequency of 10kHz and a resolution of 16 bits, capable of capturing minute changes and transient characteristics of the excitation current. Displacement data is acquired using an eddy current displacement sensor with a measurement range of ±1mm and an accuracy of 1μm, monitoring the rotor's levitation position in the magnetic field in real time. Vibration data is acquired using an accelerometer with a frequency range of 0.1Hz-5kHz and a dynamic range greater than 80dB, monitoring the rotor's vibration state and the stability of the bearing system. The PLC configuration information includes core control system information such as control strategy parameters, protection threshold settings, operating mode configuration, and communication protocol parameters. The compressor's operating data includes thermodynamic parameters such as suction pressure, discharge pressure, suction temperature, discharge temperature, refrigerant flow rate, and power consumption. All data is recorded synchronously using timestamps to ensure time consistency across multiple data sources. Data storage employs a real-time database format, supporting high-speed data writing and fast query access.

[0030] In some embodiments, the step of generating a dynamic signal matrix by adaptively adjusting the sampling rate of the operating data includes: performing excitation-displacement coupling analysis on the operating data to obtain coupling characteristics; identifying magnetic nonlinear intervals based on the coupling characteristics to generate a nonlinear identifier sequence; adaptively adjusting the sampling density according to the nonlinear identifier sequence to determine variable-density sampling points; and reconstructing a dynamic signal matrix using the data from the variable-density sampling points.

[0031] Excitation-displacement coupling analysis is performed on the operating data to obtain coupling characteristics. The excitation current signal and displacement signal in the collected operating data are analyzed to obtain the time-domain and frequency-domain correlation between changes in excitation current and rotor displacement. In radial magnetic bearings, an increase in excitation current generates a stronger magnetic force, causing the rotor to move towards the magnetic poles, forming a negative feedback control relationship. The extraction of coupling characteristics focuses on the transfer function characteristics between excitation current and displacement, including parameters such as gain coefficient, phase delay, and frequency response. The gain coefficient reflects the sensitivity of the excitation current to displacement control; a larger value indicates a faster control response. The phase delay reflects the dynamic response characteristics of the control system; excessive delay can affect system stability. The frequency response characteristics reflect the system's ability to suppress disturbances at different frequencies; systems with good high-frequency responses have strong anti-interference capabilities. A sliding window analysis method is used to extract the time-varying characteristics of the coupling features. The window length is set to 100 milliseconds, and the sliding step size is 10 milliseconds, which can track the dynamic changes in the coupling relationship. The coupling features also include the identification of nonlinear components. When the excitation current is large, magnetic saturation occurs, causing the current-displacement relationship to deviate from its linear characteristics.

[0032] The nonlinearity intervals of magnetic force are identified based on coupling characteristics, generating a nonlinear identifier sequence. Identification of these intervals involves comparing actual coupling characteristics with an ideal linear model and calculating the degree of deviation to determine the nonlinearity intensity. When the excitation current approaches the core saturation current, the magnetic field strength increases slowly, leading to a nonlinear current-magnetic force relationship. When the rotor displacement approaches the protective gap, the magnetic reluctance changes drastically, resulting in a strongly nonlinear displacement-magnetic force relationship. The nonlinearity intensity is quantified using a linearity index, defined as the ratio of the maximum deviation between the actual characteristic curve and the fitted straight line to the full scale. A linearity index less than 5% is marked as a linear interval, between 5% and 15% as a weak nonlinear interval, and greater than 15% as a strong nonlinear interval. The nonlinear identifier sequence is generated using a time-series labeling method, dividing the operating time into fixed-length time slices, each corresponding to a nonlinear identifier. The identifier sequence includes three types: linear identifier "0", weak nonlinear identifier "1", and strong nonlinear identifier "2". By statistically analyzing the distribution characteristics of the nonlinear identifier sequence, the operating state mode and nonlinear behavior patterns of the magnetic levitation system can be identified.

[0033] Variable-density sampling points are determined through adaptive adjustment of sampling density based on the nonlinear identifier sequence. A regional sampling strategy is employed to achieve adaptive adjustment of sampling density, assigning different sampling densities according to the identifier type of the nonlinear identifier sequence. The linear interval corresponds to identifier "0," where signal changes are highly regular and predictable; a low-density sampling strategy is used, with the sampling frequency set to 0.5 times the base frequency. The weakly nonlinear interval corresponds to identifier "1," where signal characteristics are between linear and strongly nonlinear; a medium-density sampling strategy is used, with the sampling frequency set to 1.0 times the base frequency. The strongly nonlinear interval corresponds to identifier "2," where signal changes are complex and difficult to predict; a high-density sampling strategy is used, with the sampling frequency set to 2.0 times the base frequency. An adaptive interpolation method is used to determine the variable-density sampling points, ensuring a smooth transition at the density change boundaries and avoiding abrupt changes in sampling density that could affect signal processing. When the system transitions from the linear interval to the strongly nonlinear interval, the sampling density gradually increases from low to high density, with a transition time set to 50 milliseconds. The temporal distribution of sampling points uses a non-uniform interval, with smaller intervals in the high-density interval and larger intervals in the low-density interval. By employing a variable density sampling strategy, data acquisition efficiency was optimized while ensuring the integrity of signal capture.

[0034] A dynamic signal matrix is ​​constructed using data from variable-density sampling points. Data reconstruction is performed using a time-base alignment method, with fixed time intervals serving as the column bases of the matrix, mapping sampling data of different densities onto a unified time grid. For high-density sampling intervals, the sampled values ​​at the corresponding time points are directly selected and filled into the matrix. For low-density sampling intervals, linear interpolation is used to calculate the values ​​at missing time points. When the interval between adjacent sampling points is large, cubic spline interpolation is used to improve reconstruction accuracy and ensure the continuity and smoothness of the interpolation curve. The dynamic signal matrix is ​​structured as a multi-dimensional array. The first dimension corresponds to the signal type (excitation current, radial displacement, axial displacement, etc.), the second dimension corresponds to the time series, and the third dimension corresponds to the sampling channel. Matrix elements contain information such as signal values, timestamps, and sampling density identifiers. The reconstruction process considers the physical constraints of the signal; the excitation current cannot be negative, and the displacement signal cannot exceed physical boundaries. Continuity checks are performed on the reconstructed data to ensure that the numerical changes at adjacent time points conform to physical laws.

[0035] Step S120: Identify the key pressure ratio control loop through the dynamic signal matrix, extract the vibration envelope and surge precursor frequency from the key pressure ratio control loop to generate early warning characteristic parameters, and use the early warning characteristic parameters to reverse adjust the stiffness damping ratio of the magnetic bearing to determine the anti-surge protection boundary.

[0036] Specifically, key pressure ratio control loops are identified using a dynamic signal matrix. The dynamic signal matrix is ​​used as the analysis input, and pressure ratio control loop identification technology is employed to determine the most critical control element affecting the surge characteristics of the heat pump. When the heat pump system is running, the compressor speed controller receives the pressure ratio setpoint and controls the compressor speed by adjusting the motor frequency. Simultaneously, the outlet throttle valve adjusts its opening based on pressure feedback. These two elements constitute the main pressure ratio control loop. The identification of key pressure ratio control loops uses signal correlation analysis to calculate the correlation strength between each signal channel in the dynamic signal matrix and pressure ratio changes. The pressure ratio control loop includes pressure sensors, PID controllers, solenoid valves, and compressor speed control, each contributing differently to pressure ratio control. By analyzing the coupling relationship between compressor inlet pressure, outlet pressure, speed signals, and valve opening signals, the key signal paths constituting the closed-loop control are identified. The determination of key control loops uses control gain analysis to calculate the sensitivity of each control element to pressure ratio adjustment; elements with high sensitivity are identified as key control loops. The main pressure ratio control loop typically includes two parallel channels: compressor speed control and outlet throttle valve control. The auxiliary control loop includes adjustment mechanisms such as bypass valve control and inlet guide vane control. Through timing analysis of the dynamic signal matrix, the activation modes and switching logic of the control loop under different operating conditions are identified.

[0037] In some embodiments, the step of extracting the vibration envelope and surge precursor frequency from the critical pressure ratio control loop to generate early warning feature parameters includes: establishing a pressure ratio-flow characteristic mapping based on the critical pressure ratio control loop; extracting vibration signals from the pressure ratio-flow characteristic mapping to form a vibration envelope; identifying frequency abrupt change points along the peak value variation of the vibration envelope; and using the recurrence period of the frequency abrupt change points as surge precursor frequencies to form early warning feature parameters.

[0038] A pressure ratio-flow characteristic mapping is established based on the key pressure ratio control loop. Using the identified key pressure ratio control loop signals, a pressure ratio-flow relationship mapping reflecting the heat pump's operating characteristics is constructed through characteristic mapping technology. When the heat pump operates under rated conditions, the pressure ratio is 3.5 and the refrigerant flow rate is 200 kg / h. This operating point is located at the center of the stable operating region in the pressure ratio-flow rate coordinate system. When the load increases, causing the flow rate to rise to 250 kg / h, the pressure ratio correspondingly decreases to 3.2, and the operating point shifts to the lower right. The establishment of the pressure ratio-flow rate characteristic mapping adopts a data-driven modeling method, collecting pressure ratio and flow rate data under different operating conditions, and establishing a quantitative relationship between the two through statistical fitting. The pressure ratio is defined as the ratio of the compressor outlet pressure to the inlet pressure, and the flow rate is expressed as the refrigerant mass flow rate. Under normal operating conditions, the pressure ratio and flow rate exhibit a stable inverse relationship; as the flow rate increases, the pressure ratio decreases accordingly. The characteristic mapping is constructed using a two-dimensional coordinate system, with the horizontal axis representing the refrigerant flow rate and the vertical axis representing the pressure ratio. The trajectory of the operating point in the coordinate system reflects the heat pump's operating state. The surge boundary is represented by a specific curve in the pressure ratio-flow rate diagram, with the unstable operating region on the left and the stable operating region on the right. When the operating point approaches the surge boundary, the pressure ratio-flow rate relationship exhibits nonlinear characteristics, and the normal inverse relationship is broken. The accuracy of the mapping relationship is determined through regression analysis of multi-condition data; the regression coefficient reflects the strength of the relationship, and the correlation coefficient reflects the tightness of the relationship.

[0039] Vibration signals are extracted from the pressure-flow ratio characteristic mapping to form a vibration envelope. Based on the established pressure-flow ratio characteristic mapping, vibration information reflecting the dynamic characteristics of the system is obtained and processed by envelope analysis. When the heat pump is operating normally, the operating point moves smoothly along the ideal pressure-flow ratio curve, and the system vibration is small. When pressure pulsation occurs, the operating point deviates from the ideal curve and oscillates. The degree of deviation reflects the vibration intensity; the larger the deviation, the more severe the vibration. Vibration signal extraction is achieved by analyzing the deviation of the mapping, calculating the degree of deviation between the actual operating point and the ideal characteristic mapping as an indicator of vibration intensity. The deviation calculation formula is D=(π_actual-π_ideal)²+(Q_actual-Q_ideal)², and the vibration intensity I=D / D_max, where π_actual is the actual pressure ratio, π_ideal is the ideal pressure ratio, Q_actual is the actual flow rate, Q_ideal is the ideal flow rate, I is the vibration intensity (a dimensionless vibration index after normalization), and D_max is the maximum deviation, referring to the maximum deviation value that occurs under all possible operating conditions, used as a normalization benchmark. When the operating point operates strictly according to the characteristic mapping, the deviation D is zero, and the system is in a stable state. When disturbances such as pressure pulsations and flow fluctuations occur, the operating point deviates from the ideal mapping trajectory. An increase in deviation indicates intensified vibration, and the normalized vibration intensity I increases accordingly. The vibration envelope is formed using a sliding window envelope detection method, which finds the maximum deviation within a time window as the envelope point. The window length is set to 5 seconds, and the sliding step size is set to 0.5 seconds, which can track the dynamic changes of the vibration envelope. The envelope signal reflects the slow trend of vibration intensity, and the envelope peak corresponds to the moment of strongest vibration. When the heat pump approaches an unstable operating condition, the operating point shows an oscillating trajectory in the pressure ratio-flow rate diagram, and the corresponding vibration envelope exhibits periodic fluctuation characteristics.

[0040] For example, identifying frequency abrupt change points along the peak value variation of the vibration envelope includes: acquiring the peak value sequence of the vibration envelope and extracting the fundamental frequency component; performing wavelet decomposition on the fundamental frequency component to identify subharmonic components; capturing non-integer multiple frequency features from the subharmonic components; and marking frequency abrupt change points according to the occurrence time of the non-integer multiple frequency features.

[0041] The peak sequence of the vibration envelope is obtained and the fundamental frequency component is extracted. All local peak points are extracted from the generated vibration envelope signal to form a peak sequence, and the fundamental frequency component is extracted from the sequence through spectral analysis. The peak sequence is obtained by detecting local extrema, searching for all peak points in the vibration envelope signal whose values ​​are greater than their adjacent points. The peak detection window length is set to 20 sampling points to ensure that spurious peaks caused by noise are filtered out. When the heat pump is operating under stable conditions, the peaks of the vibration envelope are mainly caused by the unbalanced excitation of the rotor, and the peak frequency is equal to the rotational frequency. When the system approaches surge, the unstable excitation of the fluid will generate additional peaks in the envelope, making the peak sequence more complex. The fundamental frequency component is extracted using the Fast Fourier Transform method, and the time intervals of the peak sequence are analyzed spectrally. The fundamental frequency component corresponds to the most important periodic component in the peak sequence and is usually related to the compressor's rotational frequency. Under normal operating conditions, the frequency of the fundamental frequency component is close to the rotational frequency, and its amplitude is dominant. The identification of the fundamental frequency component is achieved through a spectral peak search method, selecting the frequency component with the largest amplitude in the spectrum as the fundamental frequency.

[0042] Subharmonic components are identified by wavelet decomposition of the fundamental frequency component. The Morlet wavelet is used as the mother wavelet function, which possesses good time-frequency localization characteristics, making it suitable for analyzing subharmonic components in non-stationary signals. The decomposition process projects the fundamental frequency component onto wavelet functions of different scales to obtain component information within different frequency ranges. By adjusting the wavelet scale parameters, subharmonic frequency bands such as 1 / 2, 1 / 3, and 1 / 4 of the fundamental frequency can be analyzed. Subharmonic component identification employs wavelet coefficient amplitude analysis; a significant increase in the amplitude of wavelet coefficients at a certain scale indicates an enhancement of the corresponding frequency's subharmonic component. Before surge occurs, the fluid inside the compressor generates subharmonic oscillations, manifesting as 1 / 2 and 1 / 3 harmonics in the fundamental frequency component. The intensity of the subharmonic components is quantified by calculating the energy of the wavelet coefficients; higher energy indicates a more pronounced subharmonic component. The temporal resolution of wavelet decomposition can capture the intermittent occurrence of subharmonic components; when subharmonic components appear continuously, it indicates that the system has entered an unstable state.

[0043] This study captures non-integer multiple frequency features from subharmonic components. By analyzing the identified subharmonic components, the focus is on capturing non-integer multiple frequency features, which are often closely related to fluid instability phenomena. The capture of non-integer multiple frequency features employs refined spectral analysis methods to perform high-resolution spectral estimation of the subharmonic components. Traditional integer multiple frequencies (such as 1 / 2, 1 / 3, and 1 / 4 harmonics) correspond to stable subharmonic oscillations, while non-integer multiple frequencies (such as 0.37 and 0.68 harmonics) typically represent complex fluid dynamic phenomena. The occurrence of non-integer multiple frequencies is associated with phenomena such as rotary stall within compressors, frequency modulation of blades during passage, and flow channel resonance. The identification of frequency features uses parametric spectral estimation methods to improve frequency resolution and accurately capture non-integer multiple components. When stable non-integer multiple frequency components are detected, it indicates the presence of specific fluid instability modes in the system. Variations in the amplitude of non-integer multiple frequencies reflect the development of instability; an increase in amplitude indicates an intensification of instability. Frequency features also include frequency drift phenomena; slow changes in non-integer multiple frequency values ​​reflect the gradual change in system operating conditions.

[0044] Frequency abrupt changes are marked based on the occurrence of non-integer multiples of frequency characteristics. When the heat pump compressor is operating normally, the vibration spectrum mainly shows a 50Hz rotational frequency and its integer multiples, indicating stable system operation. When the system begins to approach surge, a non-integer multiple of frequency component (approximately 0.37 times the rotational frequency) suddenly appears in the spectrum; the first occurrence of this frequency is marked as a frequency abrupt change. Frequency abrupt changes are marked based on characteristic thresholds; when the amplitude of the non-integer multiple frequency characteristic exceeds a set threshold, an abrupt change is indicated. There are two types of abrupt changes: feature appearance abrupt changes and feature disappearance abrupt changes. When a new non-integer multiple frequency appears when the compressor enters an unstable operating condition, it indicates a feature appearance abrupt change; when these abnormal frequencies disappear after the system stabilizes, it indicates a feature disappearance abrupt change. Precise location of the abrupt change time is achieved using phase continuity analysis, which determines the exact timing of the abrupt change by monitoring the continuous changes in the phase of the frequency components. When a heat pump transitions from a stable operating state to a surge precursor state, the simultaneous detection of multiple non-integer multiple frequency characteristics (such as 18.5Hz and 34.2Hz) indicates that the system is undergoing significant dynamic characteristic changes.

[0045] The recurrence period of frequency abrupt change points is used as a surge precursor frequency to form an early warning characteristic parameter. Using identified frequency abrupt change point information, the periodicity of the abrupt change points is determined by analyzing the recurrence period, and the periodic frequency is used as a characteristic parameter of the surge precursor. The analysis of the recurrence period is based on the statistical analysis of the time series of abrupt change points, calculating the time interval distribution between abrupt change points of the same type. When the system approaches a surge state, the frequency abrupt change points will exhibit quasi-periodic recurrence characteristics, and the recurrence period corresponds to the characteristic frequency of the surge. The calculation of the recurrence period uses the autocorrelation analysis method, performing autocorrelation calculations on the time series of abrupt change points to find the time delay corresponding to the peak of the autocorrelation function. The time delay corresponding to the first significant peak is the recurrence period, and the reciprocal of the period is the surge precursor frequency. The surge precursor frequency is usually a fraction of the compressor rotation frequency, typically 0.1-0.3 times the rotation frequency. The generation of the early warning characteristic parameter uses a frequency feature fusion method, comprehensively analyzing the recurrence period frequency and the modulation frequency of the vibration envelope. When the recurrence period frequency matches the known surge characteristic frequency, the early warning level is increased. When multiple frequency characteristics show abnormalities simultaneously, the comprehensive early warning characteristic parameters reach the early warning threshold. The characteristic parameters also include frequency stability indicators; stable recurrence periods correspond to clear surge precursors, while unstable periods may be random disturbances.

[0046] The anti-surge protection boundary is determined by adjusting the stiffness-damping ratio of the magnetic bearing using early warning characteristic parameters. Utilizing the generated early warning characteristic parameters, the stiffness and damping characteristics of the magnetic levitation bearing are dynamically adjusted through reverse adjustment technology to establish an effective anti-surge protection boundary. When a surge warning is detected, the system adopts an adaptive adjustment strategy: the magnetic bearing stiffness is adjusted by K_new = K_base(1 + α·P_level), and the damping is adjusted by C_new = C_base(1 + β·P_level), where K_base and C_base are the base stiffness and damping, P_level is the warning level (0-1), and α and β are adjustment coefficients. For a mild warning, P_level = 0.15, α = 1.0, corresponding to a 15% increase in bearing stiffness; for a moderate warning, P_level = 0.25, α = 0.8, β = 1.2, resulting in a 20% increase in stiffness and a 30% increase in damping, achieved through proportional and differential gain adjustment of the excitation current. The reverse adjustment of the magnetic bearing stiffness-damping ratio employs an adaptive control method, adjusting the bearing's support characteristics in real time based on changes in early warning characteristic parameters. The anti-surge protection boundary is determined using a multi-level threshold setting method, with different protection actions set according to the early warning level. The first-level protection boundary corresponds to a mild early warning, triggering slight adjustments to the magnetic bearing parameters and alarm displays of operating parameters. The second-level protection boundary corresponds to a moderate early warning, triggering an active reduction in compressor speed and significant adjustments to the magnetic bearing parameters. The third-level protection boundary corresponds to a severe early warning, triggering emergency shutdown protection and system safety locking. Through continuous monitoring of early warning characteristic parameters and dynamic adjustment of magnetic bearing characteristics, an active anti-surge protection system is formed.

[0047] Step S130: Monitor the fin temperature difference detection information of the anti-surge protection boundary, perform asymmetric frosting rate analysis on the fin temperature difference detection information through PLC configuration information to generate a partition defrosting map, and perform minimum energy consumption analysis based on the partition defrosting map to determine the switching sequence of the reversing valve.

[0048] Specifically, the monitoring of fin temperature difference information within the anti-surge protection boundary is employed. A multi-point temperature sensor arrangement is used, with sensors installed on the evaporator's inlet, intermediate layer, and outlet surfaces to monitor temperature differences on the fin surfaces. The sensor density is 16 measurement points per square meter to ensure the spatial distribution characteristics of the fin temperature are captured. The temperature detection accuracy requirement is ±0.1℃, with a response time of less than 5 seconds, enabling timely reflection of dynamic changes in fin temperature. The fin area within the anti-surge protection boundary corresponds to the compressor's stable operating range; temperature difference changes in this area directly affect the heat pump's heating efficiency and system stability. When the heat pump operates near the anti-surge protection boundary, the evaporator's operating conditions become more demanding, and uneven frosting is prone to occur on the fin surface. Temperature detection information includes parameters such as fin surface temperature, ambient temperature, inter-fin temperature difference, and temperature change rate. Fin surface temperature reflects the heat exchanger's operating state, ambient temperature affects the initiation conditions of frosting, inter-fin temperature difference reflects the uniformity of heat exchange, and temperature change rate reflects the dynamic characteristics of the frosting process. The detection data is acquired in real time, with a sampling frequency of 1Hz to ensure that the continuous change process of fin temperature can be tracked.

[0049] In some embodiments, the step of generating a zoned defrosting map by performing asymmetric frosting rate analysis on the fin temperature difference detection information using the PLC configuration information includes: extracting the temperature gradient distribution from the fin temperature difference detection information; mapping the temperature gradient distribution to the sensor positions in the PLC configuration information to form a temperature-position correlation matrix; identifying the frosting unevenness and determining the defrosting zones based on the temperature-position correlation matrix; and spatially encoding the defrosting zones to form a zoned defrosting map.

[0050] Temperature gradient distribution is extracted from fin temperature difference detection information. The extraction employs a spatial difference method, calculating the ratio of temperature difference to distance between adjacent measuring points to obtain the magnitude of the temperature gradient in each direction. In the horizontal direction of the evaporator, the temperature gradient from the inlet to the outlet reflects the uniformity of heat transfer; areas with large gradients indicate uneven heat transfer. In the vertical direction, the temperature gradient from top to bottom reflects the influence of gravity and natural convection on heat transfer. The gradient distribution is calculated using a two-dimensional gradient operator, calculating the temperature gradient components in the x and y directions respectively, and obtaining the magnitude and direction of the gradient through vector synthesis. Areas with large temperature gradients typically correspond to locations with high frosting rates, as areas with drastic temperature changes are prone to supercooling. The gradient distribution also reflects the heat transfer characteristics of the fin surface; areas with uniform gradients indicate stable heat transfer, while areas with large gradient variations indicate unstable heat transfer. Statistical analysis of the spatial characteristics of the gradient distribution identifies areas in the evaporator with significant differences in heat transfer performance. The extracted temperature gradient distribution is represented as a two-dimensional matrix, where matrix elements correspond to the gradient values ​​at each measuring point, and the rows and columns of the matrix correspond to the spatial coordinates of the evaporator.

[0051] A temperature-position correlation matrix is ​​formed by mapping the sensor positions in the PLC configuration information to the temperature gradient distribution. Through coordinate transformation, the sensor numbers and logical addresses in the PLC configuration are converted into the physical coordinates of the evaporator. The PLC configuration information includes detailed parameters such as the installation location, measurement range, correction coefficient, and communication address of each sensor. The position mapping process establishes a correspondence between sensor numbers and spatial coordinates; sensor number T01 corresponds to the upper left corner of the evaporator (0,0), and sensor number T16 corresponds to the lower right corner (400,300). The temperature-position correlation matrix is ​​constructed using an interpolation fitting method. Based on the known temperature gradients at the measurement points, the gradient estimates for unmeasured points are calculated through spatial interpolation. The rows of the correlation matrix correspond to the y-coordinates of the evaporator, the columns correspond to the x-coordinates, and the matrix elements are the temperature gradient values ​​at the corresponding locations. The spatial resolution of the matrix is ​​set to 10mm × 10mm, providing detailed temperature distribution information. The correlation matrix visually displays the temperature gradient distribution on the evaporator surface; the color intensity reflects the gradient magnitude, and the color distribution reflects the spatial variation of the gradient.

[0052] The defrosting zones are determined based on the temperature-location correlation matrix to identify frost unevenness. Based on the established temperature-location correlation matrix, the uneven distribution of frost on the evaporator surface is identified by analyzing frost unevenness, thus determining a reasonable defrosting zoning scheme. Regions with similar temperature gradients are grouped into the same category, while regions with large gradient differences are divided into different categories to identify frost unevenness. The coefficient of variation method is used to quantify unevenness, calculating the ratio of the standard deviation to the mean of the temperature gradient in each region. Regions with a larger ratio have higher unevenness, and regions with a smaller ratio have lower unevenness. The defrosting zones are determined using a watershed segmentation method, with the extreme points of the temperature gradient as the boundaries of the zones, dividing the correlation matrix into several connected regions. The temperature gradient distribution within each zone is relatively uniform, but there are significant differences in temperature gradients between zones. The zoning scheme considers the severity of frost and the feasibility of defrosting operations; severely and concentrated frost areas are divided into independent zones, while lightly and dispersed frost areas are merged into comprehensive zones. The size of the zones is controlled within an appropriate range; overly small zones increase control complexity, while overly large zones reduce the targeting of defrosting. A typical defrosting zone scheme divides the evaporator into 4-6 zones, with each zone having a relatively uniform area, making it easy to control the defrosting operation independently.

[0053] Spatial coding is used to create a zonal defrost map for defrost zones. Using defined defrost zone information, a unique identifier code is assigned to each zone through spatial coding technology, generating a zonal defrost map that is easy for the system to identify and control. This is achieved through hierarchical coding, including three levels: area coding, location coding, and attribute coding. Area coding uses a single letter to represent the basic identifier of the zone; zones A, B, C, etc., correspond to different defrost zones. Location coding uses numbers to represent the relative position of the zone within the evaporator: 01 for the upper left zone, 02 for the upper right zone, 03 for the lower left zone, and 04 for the lower right zone. Attribute coding uses numbers to represent the frosting characteristics of the zone: 1 for heavy frosting, 2 for moderate frosting, and 3 for light frosting. Combining these codes forms a complete zone identifier, such as A01-1 representing the heavy frosting zone in the upper left of zone A. The zonal defrost map is generated using a graphical representation method, overlaying the coded information onto a plan view of the evaporator to create an intuitive zonal distribution map. In the map, different colors represent different zones, color intensity indicates the degree of frost, and numerical labels indicate defrosting priority. The map also includes geometric information about the zones, such as spatial parameters like zone area, boundary coordinates, and center point location.

[0054] In some embodiments, determining the switching sequence of the reversing valve based on the minimum energy consumption analysis of the partitioned defrost map includes: decomposing the partitioned defrost map into multi-level defrost regions according to the frosting gradient; performing heat demand inversion on the defrost regions to obtain the defrost energy consumption distribution; determining the off-peak defrost window by reverse coupling of heating demand through the defrost energy consumption distribution; and sorting the off-peak defrost windows according to the principle of minimizing the number of four-way valve operations to form the switching sequence of the reversing valve.

[0055] The zonal defrosting map is decomposed into multi-level defrosting zones based on the frost gradient. Frost gradient analysis is performed on the generated zonal defrosting map, and grading criteria are determined based on parameters such as frost thickness, frost rate, and frost density. Level 1 defrosting zones correspond to heavily frosted areas with frost thickness greater than 5 mm and frost rate higher than 2 mm / h, requiring immediate defrosting. Level 2 defrosting zones correspond to moderately frosted areas with frost thickness between 2-5 mm and frost rate between 1-2 mm / h, requiring planned defrosting. Level 3 defrosting zones correspond to lightly frosted areas with frost thickness less than 2 mm and frost rate less than 1 mm / h, allowing for delayed defrosting. The decomposition process considers the frost correlation between adjacent zones; severely frosted zones can affect the heat exchange efficiency of adjacent zones, leading to frost diffusion. The multi-level defrosting zone division also considers the economics of defrosting operations, assigning areas with high defrosting costs and limited effectiveness to lower priority zones. The results of the zone decomposition are represented in a hierarchical structure, with the upper level representing the defrosting level and the lower level representing the specific zone list. The number of zones and total area included in each level reflect the scale and complexity of the defrosting task at that level.

[0056] The defrosting energy consumption distribution is obtained by performing heat demand inversion on the defrosting area. Using decomposed multi-level defrosting area information, the heat required for defrosting in each area is calculated through heat demand inversion technology, yielding the spatial distribution characteristics of defrosting energy consumption. Heat demand inversion uses known conditions such as frost thickness, frost density, and heat exchange area to infer the heat input required for defrosting. The defrosting heat demand includes three components: latent heat of defrosting, sensible heat of fin heating, and heat loss to the environment. Latent heat of defrosting is calculated based on the frost mass and the latent heat of ice melting, with a value of 334 kJ / kg. Sensible heat of fin heating is calculated based on the fin material, mass, and temperature rise amplitude; the specific heat capacity of aluminum fins is 0.9 kJ / (kg·℃). Heat loss to the environment is calculated based on natural convection and radiation heat transfer during the defrosting process, accounting for approximately 10-15% of the total heat. The defrosting energy consumption distribution is obtained using a regional integration method. The total energy consumption is obtained by summing the heat demands of each defrosting zone, and the ratio of each zone's energy consumption to the total energy consumption is used to determine the energy consumption distribution. The energy consumption of a Level 1 defrosting zone typically accounts for 50-60%, Level 2 defrosting zones for 25-35%, and Level 3 defrosting zones for 10-20%. The energy consumption distribution also includes time-related information; differences in defrosting time demands across different zones result in uneven temporal distribution of power demand.

[0057] The peak-shifting defrosting window was determined by inverse coupling of heating demand through defrosting energy consumption distribution. The inverse coupling analysis started with the high energy consumption ratio of 50-60% in the primary defrosting area, calculating the corresponding heating capacity loss. When defrosting consumes 60% of the total energy, the remaining heating capacity is only 40%. Therefore, it is necessary to find time periods when heating demand is below 40% as the peak-shifting defrosting window for this level. Analysis of the 24-hour heating load curve shows that user heating demand drops to 30-35% of the peak value between 23:00 and 06:00 at night, matching the energy consumption impact of the primary defrosting area. The moderate energy consumption ratio of 25-35% in the secondary defrosting area corresponds to a heating capacity loss that keeps the remaining heating capacity at 65-75%. Inverse coupling analysis identified time periods when heating demand is below 65% as the matching window. During the midday period from 11:00 to 14:00, outdoor temperatures are high, and the heating load drops to 60-65% of the peak value; the heating demand during this period matches the energy consumption impact of the secondary defrosting area. The low energy consumption of 10-20% in the Level 3 defrosting zone has minimal impact on heating capacity, with the remaining heating capacity maintained at 80-90%, allowing defrosting operations to be scheduled in multiple time periods. The core of reverse coupling is establishing a quantitative relationship between defrosting energy consumption and heating loss, ensuring through energy balance analysis that the defrosting process does not lead to insufficient heating capacity. When the outdoor temperature is -5℃, the heat pump heating capacity is at its rated value. A 60% energy consumption in the Level 1 defrosting zone will reduce heating output to 40%, therefore this level can only be scheduled during late-night hours when heating demand is below 35%. When the outdoor temperature rises to 5℃, the heat pump heating load naturally decreases, providing more off-peak defrosting windows. The final determination of the off-peak defrosting window considers electricity pricing policies and grid load conditions, prioritizing off-peak hours with lower electricity prices to ensure that defrosting costs are minimized while meeting the constraint of unaffected heating demand.

[0058] The staggered defrosting windows are sorted according to the principle of minimizing the number of four-way valve operations to form the switching sequence of the reversing valve. The core of the principle of minimizing the number of four-way valve operations is to reorganize the scattered staggered defrosting windows by time, reducing the frequent switching between heating and defrosting modes. When there are 6 independent staggered defrosting windows, the traditional method requires 12 four-way valve operations (6 times from heating to defrosting, and 6 times from defrosting to heating). By adopting the principle of minimizing the number of operations, staggered defrosting windows with similar times are merged and sorted. The three windows at night of 23:30, 00:45, and 02:15 are merged into a continuous defrosting period from 23:30 to 03:00, and the three windows at noon of 11:45, 12:30, and 13:15 are merged into a continuous defrosting period from 11:45 to 14:00. Through this sorting and merging, the number of four-way valve operations is reduced from 12 to 4 (2 times from heating to defrosting, and 2 times from defrosting to heating). The sorting process prioritizes the feasibility of window merging, merging windows with time intervals of less than 45 minutes. Within the merged continuous defrosting period, a zoned sequential defrosting method is adopted. The four-way valve remains in the defrosting position, and the start and stop of defrosting in different zones are controlled only by solenoid valves, further avoiding additional actions of the four-way valve. The principle of minimum number of actions also considers ensuring defrosting effectiveness, ensuring that the merged continuous period can meet the defrosting needs of each zone. The final output of the directional valve switching sequence is in the form of a time-action matrix, clearly indicating the time, direction, and duration of each four-way valve action.

[0059] Step S140: Use the switching timing of the reversing valve and the dynamic signal matrix to perform coupling analysis of operating characteristics and oil-gas migration to generate oil-gas separation optimization parameters. Adjust the magnetic levitation gap according to the oil-gas separation optimization parameters to extract the zero-power suspension interval. Construct a hierarchical energy-saving control configuration based on the zero-power suspension interval.

[0060] Specifically, the operating characteristics and oil-gas migration coupling analysis are performed using the reversing valve switching timing and dynamic signal matrix to generate optimized oil-gas separation parameters. When the heat pump performs defrosting, the reversing valve switching changes the refrigerant flow direction and simultaneously causes a redistribution of lubricating oil in the system. Oil-gas migration affects the compressor's suspension characteristics and operating efficiency. The operating characteristics and oil-gas migration coupling analysis employs a multi-parameter correlation method, comprehensively considering factors such as the reversing valve actuation time, refrigerant flow rate changes, oil viscosity changes, and magnetic bearing load changes. The excitation current, displacement signal, and vibration signal in the dynamic signal matrix exhibit characteristic changes during the reversing valve switching process, reflecting the dynamic evolution of the oil-gas separation state. The oil-gas migration coupling analysis focuses on the migration pattern of oil between the evaporator and condenser, and the impact of this migration process on the internal oil-gas balance of the compressor. When oil flows back from the evaporator to the compressor, it changes the compressor's mass distribution and center of gravity position, affecting the control parameters of the magnetic bearing. The generation of optimized oil-gas separation parameters uses a feature extraction method to extract key parameters affecting the oil-gas separation effect from the coupling analysis results. The optimization parameters include control variables such as oil reflux rate, gas-liquid separation efficiency, oil viscosity correction coefficient, and separator pressure drop compensation.

[0061] In some embodiments, the step of adjusting the magnetic levitation gap according to the oil-gas separation optimization parameters to extract the zero-power levitation interval includes: searching for the levitation equilibrium point of the oil-gas separation optimization parameters to obtain a static equilibrium position; adjusting the magnetic levitation gap based on the static equilibrium position to form a gap optimization sequence; selecting the point with the minimum electromagnetic power consumption from the gap optimization sequence to obtain a power consumption valley value; and expanding the operating conditions of the power consumption valley value to form a zero-power levitation interval.

[0062] The static equilibrium position is obtained by searching for the suspension equilibrium point using the optimized oil-gas separation parameters. The search is conducted by comprehensively considering the balance of various forces, including electromagnetic attraction, gravity, centrifugal force, and oil damping force, using the obtained optimized oil-gas separation parameters. The force balance equation is F_mag + F_gravity + F_centrifugal + F_damping = 0, where F_mag is the electromagnetic attraction, F_gravity is gravity, F_centrifugal is centrifugal force, and F_damping is oil damping force. The oil return rate affects the rotor's center of gravity position change Δm·g; a high return rate shifts the center of gravity downwards, requiring corresponding adjustments to the attraction force of the lower magnetic bearing. The gas-liquid separation efficiency parameter affects the oil content in the bearing clearance; low efficiency results in a high oil content, increasing suspension damping and altering the equilibrium position. The static equilibrium position search is performed iteratively by solving the force balance equation, gradually adjusting radial and axial displacements from the initial position to find the equilibrium point where the resultant force in all directions is zero. The search process considers the influence of the oil viscosity correction factor. Higher viscosity oil increases rotor damping, requiring greater levitation force to overcome the damping effect. The separator pressure drop compensation parameter affects the system pressure distribution; changes in pressure drop alter the oil distribution within the bearing cavity, necessitating corresponding force compensation corrections during the equilibrium point search. The equilibrium point search range is limited to the safe levitation region to ensure the rotor does not contact the stator. The obtained static equilibrium position includes coordinate values ​​in the x, y, and z directions, along with the corresponding force balance parameters.

[0063] A gap optimization sequence is formed by adjusting the magnetic levitation gap based on the static equilibrium position. Gap adjustment involves coordinating and optimizing the radial and axial gaps, fine-tuning the air gap size in each direction based on the static equilibrium position. The adjustment range for the radial gap is set to ±0.2 mm from the equilibrium position, and the adjustment range for the axial gap is set to ±0.1 mm from the equilibrium position, ensuring the safety of the adjustment process. The gap optimization sequence is generated using a parameter scanning method, sequentially changing the gap values ​​according to a set step size and recording the system response characteristics under each gap configuration. The scanning step size is set to 0.02 mm, providing sufficient resolution to identify the optimal gap. Each gap configuration in the sequence corresponds to a set of control parameters, including radial gap value, axial gap value, excitation current requirement, and levitation stiffness. The gap adjustment process considers the coupling effect of gaps in different directions; changes in the radial gap affect the distribution of the axial levitation force, and changes in the axial gap affect the magnitude of the radial stiffness. The optimization sequence is arranged according to the order of gap adjustment, forming a continuous change sequence from the minimum gap to the maximum gap.

[0064] The minimum electromagnetic power consumption point is selected from the gap optimization sequence to obtain the power consumption valley value. By analyzing the electromagnetic power consumption characteristics of each configuration in the gap optimization sequence, the gap configuration point with the minimum power consumption is selected to obtain the power consumption valley value. The formula for calculating the total electromagnetic power consumption is P_total=P_radial+P_axial+P_control=I²R_coil+U·I_control, where P_radial is the radial magnetic bearing excitation power consumption, P_axial is the axial magnetic bearing excitation power consumption, P_control is the control processing power consumption, I is the excitation current, R_coil is the coil resistance, U is the control voltage, and I_control is the control current. The power consumption includes three components: radial magnetic bearing excitation power consumption, axial magnetic bearing excitation power consumption, and control processing power consumption. The total power consumption is calculated by measuring the current and voltage parameters under different gap configurations in real time. The power consumption selection adopts a global optimization method, traversing all configuration points in the gap optimization sequence and comparing the power consumption values ​​of each point. The minimum power consumption point usually appears near a specific gap combination. This position corresponds to the natural suspension state of the rotor, requiring the least amount of external excitation support. The determination of the power consumption valley value takes into account the local characteristics of the power consumption curve, avoiding the selection of local extrema rather than the global minimum. When multiple extrema with similar power consumption exist, the configuration with better levitation stability is selected as the power consumption valley value. The obtained power consumption valley value includes complete information such as the corresponding gap configuration, power consumption value, excitation current distribution, and levitation force characteristics. The power consumption valley value also includes the dynamic response characteristics under this configuration, such as control performance indicators like step response time, overshoot, and steady-state error.

[0065] For example, the step of extending the power consumption valley to form a zero-power floating range includes: extracting the magnetic field gradient distribution based on the power consumption valley; introducing rotor gyroscope effect correction into the magnetic field gradient distribution to obtain a dynamic equilibrium domain; performing multi-condition envelope verification on the dynamic equilibrium domain to form a stable operating range; and extending the stable operating range along the power consumption gradient decreasing direction to form a zero-power floating range.

[0066] The magnetic field gradient distribution is extracted based on the power consumption valleys. The extraction employs the finite element method (FEM) for magnetic field calculation, establishing a three-dimensional magnetic field model that includes the magnetic bearing geometry, material properties, and excitation conditions. The gap configuration corresponding to the power consumption valleys serves as the boundary condition for magnetic field calculation, determining the magnetic flux density distribution within the air gap. The magnetic field gradient is calculated using the spatial difference method, calculating the spatial rate of change of magnetic flux density in both the radial and axial directions. The radial magnetic field gradient affects the rotor's levitation stiffness in the radial direction; regions with large gradients have high levitation stiffness, while regions with small gradients have low levitation stiffness. The axial magnetic field gradient affects the rotor's positioning accuracy in the axial direction; the uniformity of the gradient distribution determines the stability of axial levitation. The magnetic field gradient distribution also incorporates nonlinear characteristics of the magnetic field; when the magnetic flux density approaches saturation, the gradient exhibits nonlinear changes. The extracted magnetic field gradient distribution is represented as a three-dimensional data field, containing spatial coordinates and corresponding gradient vector information.

[0067] A dynamic equilibrium domain is obtained by incorporating rotor gyroscopic effect correction into the magnetic field gradient distribution. This correction considers the influence of gyroscopic torque generated during high-speed rotor rotation on levitation characteristics. When the rotor operates at high speed, radial disturbances generate vertical gyroscopic torque, and axial disturbances generate radial gyroscopic torque; this coupling effect alters the rotor's equilibrium position. Gyroscopic effect correction is achieved by superimposing a gyroscopic force gradient into the magnetic field gradient distribution, resulting in a more accurate reflection of the rotor's actual levitation characteristics. The dynamic equilibrium domain is obtained using stability analysis, calculating the characteristic frequencies and damping ratios of small-amplitude rotor vibrations based on the corrected gradient distribution. A stable dynamic equilibrium domain corresponds to negative real parts for all characteristic frequencies, indicating that disturbances decay naturally. The boundary of the equilibrium domain is determined by a stability criterion; points on the boundary correspond to the critical state where the system is just stable. The obtained dynamic equilibrium domain is represented in parameter space, encompassing the stability ranges of multidimensional parameters such as radial displacement, axial displacement, and rotational speed. The equilibrium domain also includes dynamic response characteristics, such as vibration modes, natural frequencies, and damping characteristics.

[0068] A stable operating range is formed by enveloping the dynamic equilibrium domain under multiple operating conditions. Typical operating conditions include different combinations of operating parameters such as refrigerant flow rate, compression ratio, ambient temperature, and load conditions. The dynamic equilibrium domain under each operating condition will change in position and size due to different operating conditions. Envelope analysis determines the stable operating range by finding the intersection of the dynamic equilibrium domains under all operating conditions. The intersection region corresponds to the parameter range that can operate stably under all operating conditions. The formation of the stable operating range takes into account the dynamic characteristics of the operating condition switching process to ensure that the system can smoothly transition during operating condition changes. The envelope process also considers the probability distribution of operating conditions, prioritizing the stable region under high-probability operating conditions and appropriately compressing the coverage of low-probability operating conditions. The formed stable operating range is represented in the form of a multi-dimensional parameter region, including the value boundaries and constraints of each parameter. The operating range also includes the distribution information of performance indicators, such as power consumption levels at different locations, response speed, and disturbance rejection capabilities.

[0069] A zero-power floating region is formed by expanding the stable operating range along the direction of decreasing power consumption gradient. Gradient analysis is used to determine the power consumption gradient vector at each point within the stable operating range, with the gradient direction pointing towards the direction of the fastest increase in power consumption and the gradient descent direction pointing towards the direction of the fastest decrease in power consumption. The expansion process starts from the center point of the stable operating range and gradually expands the region boundary along the direction of decreasing power consumption gradient until the power consumption or stability constraints are met. The expansion constraints include a power consumption increase not exceeding 10% of the initial value and a stability margin not less than twice the safety requirement. The boundary of the zero-power floating region is determined by the intersection of multiple constraints, ensuring that any operating point within the region meets the power consumption and stability requirements. An adaptive step-size method is used during the expansion process, employing small step sizes in regions with large power consumption gradients and large step sizes in regions with small power consumption gradients. The formed zero-power floating region contains complete parameter boundary information, as well as the power consumption and stability distributions within the region. The region also includes a recommended location for the optimal operating point, corresponding to the parameter combination with the minimum power consumption and the maximum stability margin.

[0070] A hierarchical energy-saving control configuration is constructed based on the zero-power floating range. A three-layer control structure is designed: an upper layer for operating condition identification, a middle layer for parameter optimization, and a lower layer for execution control. The upper layer identifies the current operating condition type based on the heat pump's operating status, including different modes such as start-up, steady-state, variable, and shutdown. The middle layer determines the optimal control parameters based on the operating condition type and zero-power floating range information, including the excitation current setpoint, the target value of the floating gap, and the control gain parameters. The lower layer is responsible for executing specific control actions, driving the magnetic levitation bearing system to operate according to the optimized parameters. Under steady-state conditions, the system prioritizes energy-saving operation using the zero-power floating range, reducing the excitation current to the minimum level. Under variable conditions, the system dynamically adjusts the floating parameters according to load changes, ensuring a balance between energy saving and stability. The hierarchical control configuration also includes operating condition switching logic, automatically switching to the corresponding control mode when operating conditions change. Configuration parameters are stored in a hierarchical data table format for easy system retrieval and parameter loading.

[0071] Step S150: Perform magnetic bearing current ripple analysis on the operating data to predict the compressor eccentricity development trend and generate a preventive maintenance time window. Match the preventive maintenance time window with the hierarchical energy-saving control configuration to determine the adaptive control strategy by fault-tolerant degradation operation mode. Based on the adaptive control strategy, perform multi-parameter fusion optimization to generate heat pump control commands.

[0072] Specifically, magnetic bearing current ripple analysis is performed on operating data to predict the compressor eccentricity development trend and generate preventative maintenance time windows. Using the excitation current, displacement signal, and vibration signal of the magnetic levitation bearing system in the operating data, ripple characteristics of the excitation current are extracted and analyzed. The excitation current signal is decomposed into DC and AC ripple components through frequency domain decomposition, with a focus on analyzing the amplitude changes and frequency characteristics of the ripple components. Current ripple is mainly caused by air gap unevenness due to rotor eccentricity; the greater the eccentricity, the more pronounced the current ripple. Ripple analysis assesses the development state of rotor eccentricity by monitoring the amplitude changes of the rotational frequency component in the excitation current, and cross-validates the accuracy of the eccentricity using displacement signals. When the rotor experiences mass imbalance or bearing wear, the eccentricity gradually increases, and the corresponding current ripple amplitude shows an upward trend.

[0073] The prediction of compressor eccentricity development trend employs time series analysis to establish a quantitative relationship model between ripple amplitude and eccentricity. By fitting the time series variation curve of current ripple amplitude, the eccentricity development trajectory for future time periods is extrapolated. The prediction model considers the nonlinear characteristics of eccentricity development, with slow initial growth followed by accelerated growth. Eccentricity prediction also incorporates the spectral characteristics of vibration signals in operational data; the presence of subharmonic components in the vibration indicates intensified eccentricity, thus correcting the prediction accuracy of the eccentricity development rate. Preventive maintenance time windows are generated based on the eccentricity development trend prediction results. A risk threshold determination method is used: the start point of the maintenance time window is determined when the predicted eccentricity reaches 80% of the safety threshold, and the end point is determined when it reaches 90%. The length of the time window is typically 7-14 days, providing sufficient preparation time for the formulation and execution of maintenance plans.

[0074] In some embodiments, the step of matching the preventive maintenance time window with the hierarchical energy-saving control configuration to determine the adaptive control strategy through fault-tolerant degradation operation mode includes: extracting maintenance priorities based on the preventive maintenance time window; selecting a degradeable control layer from the hierarchical energy-saving control configuration; sorting the degradeable control layers according to the maintenance priorities to form a degradation sequence; and forming an adaptive control strategy through the switching logic of the degradation sequence.

[0075] Maintenance priorities are extracted based on preventative maintenance time windows. Maintenance priority indicators reflecting the urgency and importance of maintenance are extracted from the generated preventative maintenance time window information. When the preventative maintenance time window indicates that the magnetic bearing eccentricity will reach a critical value within 7 days, the corresponding bearing correction maintenance is marked as high priority; when the fin defrost device needs routine cleaning within 14 days, the corresponding defrost maintenance is marked as medium priority; and when the lubricating oil replacement cycle is 30 days away, the corresponding oil maintenance is marked as low priority. The extraction of maintenance priorities uses a multi-factor evaluation method, comprehensively considering factors such as failure probability, failure severity, maintenance difficulty, and maintenance cost. Failure probability is assessed through the prediction results of eccentricity development trends; high confidence predictions correspond to high failure probabilities. Failure severity is assessed through the degree of impact of potential failures on operation; failures affecting safety are of high severity, while failures affecting efficiency are of medium severity. Maintenance difficulty is assessed through the complexity of the maintenance work and the time required; maintenance requiring disassembly of major components is of high difficulty, while maintenance involving simple adjustments and cleaning is of low difficulty. Priority indicators are quantified using a tiered scoring method, with each factor rated on a scale of 1 to 5. A weighted sum is then used to obtain a comprehensive priority score. High priority corresponds to maintenance projects with a score greater than 4.0, medium priority to projects with a score between 2.0 and 4.0, and low priority to projects with a score less than 2.0. The extracted maintenance priorities are organized in a priority list format, including information such as maintenance project, priority score, estimated maintenance time, and resource requirements.

[0076] Select a degradable control layer from the hierarchical energy-saving control configuration. The hierarchical energy-saving control configuration comprises three layers: the operating condition identification layer, the parameter optimization layer, and the execution control layer. The operating condition identification layer is responsible for basic status monitoring and is the core layer that cannot be degraded. The parameter optimization layer handles complex energy-saving parameter calculations and can be downgraded from precise optimization to a simplified calculation mode. The precise position control in the execution control layer can be downgraded from a high-precision mode to a standard-precision mode. The selection of degradable control layers employs a functional importance analysis method, distinguishing between core and auxiliary functions, prioritizing the retention of core functions, and appropriately downgrading auxiliary functions. Degradation processing includes three methods: function simplification, accuracy reduction, and response delay. Function simplification reduces processing load by disabling complex control functions; accuracy reduction improves fault tolerance by relaxing control accuracy requirements; and response delay reduces computational load by extending the control cycle. The parameter optimization layer is downgraded through function simplification, simplifying multi-objective optimization to single-objective optimization. The execution control layer is downgraded through accuracy reduction, relaxing the position control accuracy from ±0.01mm to ±0.05mm. The selected degradable control layers form a degradable candidate list, which includes attributes such as control layer name, degradation method, impact level, and recovery conditions. The candidate list also includes changes in performance metrics after degradation, used to assess the acceptability of the degradation operation.

[0077] Degradable control layers are sorted according to maintenance priority to form a degradation sequence. Priority matching is used for sorting, and the importance of maintenance priority and control layer is correlated through analysis. High-priority maintenance items require stricter degradation measures for their corresponding control layers and are executed first in the degradation sequence; medium-priority maintenance items use moderate degradation measures and are executed in the middle of the sequence; low-priority maintenance items use slight degradation measures and are executed later in the sequence. The sorting of the degradation sequence considers the dependencies of degradation operations, as the degradation of some control layers requires adjustment from other control layers. The reversibility of degradation is also considered, prioritizing easily recoverable degradation operations to avoid irreversible functional loss. The resulting degradation sequences are arranged in execution order: the first sequence is parameter optimization layer degradation, corresponding to the high-priority requirement of magnetic bearing calibration maintenance; the second sequence is execution control layer accuracy degradation, corresponding to the medium-priority requirement of defrosting maintenance; and the third sequence is response time extension, corresponding to the low-priority requirement of oil maintenance. The sequence also includes a setting for the degradation depth, which determines the degree of degradation for each control layer based on maintenance priority. High priority corresponds to deep degradation, while low priority corresponds to shallow degradation.

[0078] An adaptive control strategy is formed through the switching logic of the degradation sequence. The switching logic is designed using a conditional triggering method. When the maintenance time window enters the warning stage, the switching logic triggers the execution of the first sequence of the degradation sequence, switching the parameter optimization layer from complex multi-objective optimization to simplified single-objective optimization. When the maintenance time window enters the preparation stage, the switching logic continues to trigger the execution of the second sequence of the degradation sequence, reducing the position accuracy of the execution control layer from ±0.01mm to ±0.05mm. When the maintenance time window enters the execution stage, the switching logic triggers the execution of the third sequence of the degradation sequence, extending the control response cycle from 10ms to 50ms. The switching logic of the degradation sequence adopts a sequential triggering mechanism; the next sequence can only be triggered after the previous sequence has been completed, ensuring the orderliness and controllability of the degradation process. The switching logic also includes a sequence rollback mechanism; after maintenance is completed, the switching logic restores the normal functions of each control layer in reverse sequence. The state transition conditions of the switching logic include trigger signals such as maintenance time nodes, changes in equipment health status, and external maintenance command inputs. The adaptive control strategy achieves a smooth transition from normal operation mode to maintenance mode through the switching logic of the degradation sequence, ensuring the stable operation of basic functions during maintenance.

[0079] The heat pump control commands are generated through multi-parameter fusion optimization based on an adaptive control strategy. When the adaptive control strategy instructs the execution of the first sequence of degradation, the multi-parameter fusion process converts the degradation requirements of the parameter optimization layer into simplified control parameters for the magnetic levitation bearing, while simultaneously adjusting the compressor speed limit and load distribution ratio. When the strategy enters the second sequence of degradation, the fusion optimization converts the precision degradation requirements of the execution control layer into parameter adjustment commands for the position controller and correction commands for the reversing valve action timing. The multi-parameter fusion optimization employs a constraint-solving method, comprehensively considering the control requirements of multiple subsystems, including magnetic levitation bearing control, compressor operation control, reversing valve operation control, and defrosting timing control. The parameter fusion process considers the coupling relationships and constraints between subsystems to avoid conflicts and interference between control commands. The objective function of the fusion optimization includes multiple optimization objectives such as operational stability, maintenance safety, and energy consumption control, which are weighted and summed to form a comprehensive objective function. The generation of heat pump control commands adopts a hierarchical command structure: upper-level commands correspond to operating modes and condition settings, middle-level commands correspond to the coordinated control parameters of each subsystem, and lower-level commands correspond to the action commands of specific actuators. The control commands contain complete information such as command type, target value, execution sequence and priority, ensuring that the heat pump can operate stably according to the adaptive control strategy.

[0080] To implement the PLC-based high-pressure ratio magnetic levitation centrifugal heat pump control method corresponding to the above method embodiments, and to achieve the corresponding functions and technical effects. See also Figure 2 , Figure 2 A structural block diagram of a PLC-based high-pressure ratio magnetic levitation centrifugal heat pump control system 200 provided in this application embodiment is shown. For ease of explanation, only the parts relevant to this embodiment are shown. The PLC-based high-pressure ratio magnetic levitation centrifugal heat pump control system 200 provided in this application embodiment includes:

[0081] The data acquisition module 201 is used to acquire the operating data and PLC configuration information of the high-pressure ratio magnetic levitation centrifugal heat pump, and to generate a dynamic signal matrix by adaptively adjusting the sampling rate of the operating data.

[0082] The anti-surge control module 202 is used to identify the key pressure ratio control loop through the dynamic signal matrix, extract the vibration envelope and surge precursor frequency from the key pressure ratio control loop to generate early warning characteristic parameters, and use the early warning characteristic parameters to reversely adjust the stiffness damping ratio of the magnetic bearing to determine the anti-surge protection boundary.

[0083] The timing switching module 203 is used to monitor the fin temperature difference detection information of the anti-surge protection boundary, perform asymmetric frosting rate analysis on the fin temperature difference detection information through the PLC configuration information to generate a partition defrosting map, and perform minimum energy consumption analysis based on the partition defrosting map to determine the switching timing of the reversing valve.

[0084] Energy-saving optimization module 204 is used to perform coupling analysis of operating characteristics and oil-gas migration using the switching timing of the reversing valve and the dynamic signal matrix to generate oil-gas separation optimization parameters, adjust the magnetic levitation gap according to the oil-gas separation optimization parameters to extract the zero-power suspension interval, and construct a hierarchical energy-saving control configuration based on the zero-power suspension interval.

[0085] The predictive control module 205 is used to perform magnetic bearing current ripple analysis on the operating data to predict the compressor eccentricity development trend and generate a preventive maintenance time window. It then matches the preventive maintenance time window with the hierarchical energy-saving control configuration to determine an adaptive control strategy by fault-tolerant degradation operation mode matching. Finally, it performs multi-parameter fusion optimization based on the adaptive control strategy to generate heat pump control commands.

[0086] The PLC-based high-pressure ratio magnetic levitation centrifugal heat pump control system 200 described above can implement the PLC-based high-pressure ratio magnetic levitation centrifugal heat pump control method of the above method embodiments. The options in the above method embodiments are also applicable to this embodiment, and will not be detailed here. The remaining contents of this application embodiment can be referred to the contents of the above method embodiments, and will not be repeated in this embodiment.

[0087] The purpose of the above embodiments is to reproduce and derive the technical solution of the present invention by way of example, and to fully describe the technical solution, purpose and effect of the present invention. The purpose is to enable the public to have a more thorough and comprehensive understanding of the disclosure of the present invention, and not to limit the scope of protection of the present invention.

[0088] The above embodiments are not an exhaustive list based on the present invention, and there may be many other embodiments not listed. Any substitutions and improvements made without departing from the concept of the present invention are within the protection scope of the present invention.

Claims

1. A PLC-based control method for a high-pressure ratio magnetic levitation centrifugal heat pump, characterized in that, include: The operation data and PLC configuration information of the high-pressure ratio magnetic levitation centrifugal heat pump are collected, and the operation data is adaptively sampled to generate a dynamic signal matrix. The key pressure ratio control loop is identified by the dynamic signal matrix. Vibration envelope and surge precursor frequency are extracted from the key pressure ratio control loop to generate early warning characteristic parameters. The early warning characteristic parameters are used to adjust the stiffness damping ratio of the magnetic bearing to determine the anti-surge protection boundary. The temperature difference detection information of the fins of the anti-surge protection boundary is monitored. The PLC configuration information is used to perform asymmetric frosting rate analysis on the temperature difference detection information of the fins to generate a partition defrosting map. Based on the partition defrosting map, minimum energy consumption analysis is performed to determine the switching sequence of the reversing valve. The switching timing of the directional valve and the dynamic signal matrix are used to perform coupling analysis of operating characteristics and oil-gas migration to generate oil-gas separation optimization parameters. The magnetic levitation gap is adjusted according to the oil-gas separation optimization parameters to extract the zero-power suspension interval. A hierarchical energy-saving control configuration is constructed based on the zero-power suspension interval. The operating data is analyzed for magnetic bearing current ripple to predict the compressor eccentricity trend and generate a preventive maintenance time window. The preventive maintenance time window is matched with the hierarchical energy-saving control configuration to determine the adaptive control strategy by fault-tolerant and degraded operation mode. Based on the adaptive control strategy, multi-parameter fusion optimization is performed to generate heat pump control commands.

2. The method according to claim 1, characterized in that, The step of adaptively adjusting the sampling rate of the running data to generate a dynamic signal matrix includes: Excitation-displacement coupling analysis was performed on the operating data to obtain coupling characteristics; Based on the aforementioned coupling features, a nonlinear identifier sequence is generated by identifying the magnetic nonlinear interval. The variable density sampling points are determined by adaptively adjusting the sampling density based on the nonlinear identifier sequence. A dynamic signal matrix is ​​formed by reconstructing the data from the variable density sampling points.

3. The method according to claim 1, characterized in that, The step of extracting vibration envelope and surge precursor frequency from the key pressure ratio control loop to generate early warning feature parameters includes: A pressure ratio-flow characteristic mapping is established based on the aforementioned key pressure ratio control loop; Vibration signals are extracted from the pressure ratio-flow characteristic mapping to form a vibration envelope; Identify frequency abrupt changes along the peak value variation of the vibration envelope; The recurrence period of the frequency abrupt change point is used as a characteristic parameter for the formation of surge precursor frequencies and early warning features.

4. The method according to claim 1, characterized in that, The step of performing minimum energy consumption analysis based on the partition defrosting map to determine the switching sequence of the reversing valves includes: The zonal defrosting map is decomposed into multi-level defrosting zones according to the frost gradient. The defrosting energy consumption distribution is obtained by performing heat demand inversion on the defrosting area. The defrosting energy consumption distribution is used to reverse couple heating demand to determine the off-peak defrosting window; The staggered defrosting windows are sorted according to the principle of minimizing the number of four-way valve operations to form the switching sequence of the reversing valve.

5. The method according to claim 1, characterized in that, The step of generating a zoned defrosting map by performing asymmetric frosting rate analysis on the fin temperature difference detection information using the PLC configuration information includes: Extract the temperature gradient distribution from the fin temperature difference detection information; The temperature gradient distribution is mapped using the sensor locations in the PLC configuration information to form a temperature-location correlation matrix; Based on the temperature-location correlation matrix, the unevenness of frost formation is identified to determine the defrosting zones; Spatial coding is performed on the defrosting zones to form a zone defrosting map.

6. The method according to claim 1, characterized in that, The step of adjusting the magnetic levitation gap to extract the zero-power levitation zone based on the oil-gas separation optimization parameters includes: The static equilibrium position is obtained by searching for the suspension equilibrium point of the oil-gas separation optimization parameters; Based on the static equilibrium position, the magnetic levitation gap is adjusted to form a gap optimization sequence; The power consumption valley value is obtained by selecting the point with the minimum electromagnetic power consumption from the gap optimization sequence; The power consumption valley value is extended by operating conditions to form a zero-power floating range.

7. The method according to claim 1, characterized in that, The step of matching the preventive maintenance time window with the hierarchical energy-saving control configuration to determine the adaptive control strategy through fault-tolerant degradation operation mode matching includes: Maintenance priorities are extracted based on the aforementioned preventative maintenance time window; Select a degradeable control layer from the hierarchical energy-saving control configuration; The degradeable control layers are sorted according to the maintenance priority to form a degrade sequence; An adaptive control strategy is formed through the switching logic of the degradation sequence.

8. The method according to claim 3, characterized in that, The identification of frequency abrupt change points by the peak value variation along the vibration envelope includes: Obtain the peak sequence of the vibration envelope and extract the fundamental frequency component; Wavelet decomposition is performed on the fundamental frequency component to identify subharmonic components; Capture non-integer multiple frequency features from the subharmonic components; Frequency abrupt change points are marked based on the occurrence time of the non-integer multiple frequency characteristics.

9. The method according to claim 6, characterized in that, The step of extending the power consumption valley to form a zero-power floating range includes: Extract the magnetic field gradient distribution based on the power consumption valley value; A dynamic equilibrium domain is obtained by introducing a rotor gyroscope effect correction into the magnetic field gradient distribution. The dynamic equilibrium domain is subjected to multi-condition envelope verification to form a stable operating range; The stable operating range is extended along the power consumption gradient decreasing direction to form a zero-power floating range.

10. A PLC-based control system for a high-pressure magnetic levitation centrifugal heat pump, characterized in that: include: The data acquisition module is used to collect the operating data and PLC configuration information of the high-pressure ratio magnetic levitation centrifugal heat pump, and to generate a dynamic signal matrix by adaptively adjusting the sampling rate of the operating data. The anti-surge control module is used to identify the key pressure ratio control loop through the dynamic signal matrix, extract the vibration envelope and surge precursor frequency from the key pressure ratio control loop to generate early warning characteristic parameters, and use the early warning characteristic parameters to reversely adjust the stiffness damping ratio of the magnetic bearing to determine the anti-surge protection boundary. The timing switching module is used to monitor the fin temperature difference detection information of the anti-surge protection boundary, perform asymmetric frosting rate analysis on the fin temperature difference detection information through the PLC configuration information to generate a partition defrosting map, and perform minimum energy consumption analysis based on the partition defrosting map to determine the switching timing of the reversing valve. The energy-saving optimization module is used to perform coupling analysis of operating characteristics and oil-gas migration using the switching timing of the reversing valve and the dynamic signal matrix to generate oil-gas separation optimization parameters, adjust the magnetic levitation gap according to the oil-gas separation optimization parameters to extract the zero-power suspension interval, and construct a hierarchical energy-saving control configuration based on the zero-power suspension interval. The predictive control module is used to perform magnetic bearing current ripple analysis on the operating data to predict the compressor eccentricity development trend and generate a preventive maintenance time window. It then matches the preventive maintenance time window with the hierarchical energy-saving control configuration to determine an adaptive control strategy by fault-tolerant degradation operation mode matching. Finally, it performs multi-parameter fusion optimization based on the adaptive control strategy to generate heat pump control commands.

Images