Power control method and system for X-ray generating device

Abstract

The present invention relates to the technical field of power supply control, in particular to a power supply control method and system for an X-ray generating device. The method includes: detecting in real time the working state parameters of the X-ray tube, and generating a corresponding voltage / current deviation correction amount according to a preset voltage-current characteristic curve; dynamically adjusting the driving frequency of the high-frequency inverter based on the voltage / current deviation correction amount, and at the same time collecting the ripple characteristic quantity fed back from the power supply output end to form a closed-loop correction factor; coupling and analyzing the closed-loop correction factor with the change rate of the cathode emission current of the X-ray tube; reconstructing the trigger timing map of the power switch device according to the phase compensation instruction, and synchronously injecting the thermodynamic parameters of the tube body for secondary duty cycle correction; generating a multi-parameter collaborative control instruction by iteratively optimizing the matching relationship between the trigger timing map and the duty cycle correction value. The adaptive and precise adjustment of the power supply system is realized, so that the X-ray tube can obtain optimized power supply under different working conditions.

Description

Technical Field

[0001] The present invention relates to the technical field of power control, and particularly to a power control method and system for an X-ray generating device. Background Art

[0002] With the wide application of X-ray imaging technology, higher requirements are put forward for the power control accuracy and dynamic response ability of X-ray generating devices. Traditional power control methods mostly adopt an open-loop voltage and current regulation mechanism, which has technical bottlenecks such as response lag and low parameter matching degree. Especially under transient working conditions, it is easy to cause fluctuations in the cathode emission current of the X-ray tube, resulting in an increase in tube voltage ripple and a decrease in focus stability. Although existing technologies have tried to introduce PID feedback regulation, they have not effectively solved the non-linear coupling problem between the driving characteristics of high-frequency inverters and the dynamic characteristics of loads. Moreover, during long-term operation, the thermionic emission effect caused by the temperature rise of power devices will significantly change the equivalent impedance characteristics of the X-ray tube, causing the preset working curve to shift. In addition, conventional pulse width modulation strategies often ignore the correlation between the ripple spectrum characteristics and the switching timing, resulting in limited electromagnetic interference suppression effect. These problems jointly restrict the precise control of the X-ray output dose, affect the imaging resolution and accelerate the aging of the tube body. Therefore, it is urgent to construct an intelligent control system that integrates real-time interaction of multi-physical field parameters, and realizes the deep adaptation of power supply characteristics and the working state of the X-ray tube through a dynamic compensation mechanism, so as to improve the system energy efficiency ratio and output stability. Summary of the Invention

[0003] The present invention overcomes the deficiencies of the prior art and provides a power control method and system for an X-ray generating device.

[0004] The technical solution adopted by the present invention to achieve the above purpose is as follows:

[0005] The first aspect of the present invention discloses a power control method for an X-ray generating device, including the following steps:

[0006] Real-time detect the working state parameters of the X-ray tube, and generate corresponding voltage / current deviation correction amounts according to the preset voltage-current characteristic curve;

[0007] Based on the voltage / current deviation correction amounts, dynamically adjust the driving frequency of the high-frequency inverter, and at the same time collect the ripple characteristic quantities fed back from the power output end to form a closed-loop correction factor;

[0008] Couple and analyze the closed-loop correction factor with the change rate of the cathode emission current of the X-ray tube to generate a phase compensation instruction for the pulse width modulation signal;

[0009] According to the phase compensation instruction, reconstruct the trigger timing map of the power switch device, and synchronously inject the tube body thermodynamic parameters for secondary duty cycle correction;

[0010] Generate multi-parameter collaborative control instructions by iteratively optimizing the matching relationship between the trigger timing map and the duty cycle correction value.

[0011] Preferably, the working state parameters of the X-ray tube are detected in real time, and the corresponding voltage / current deviation correction amount is generated according to the preset voltage-current characteristic curve. Specifically:

[0012] Parallel data acquisition is performed on the anode voltage and cathode current of the X-ray tube, and high-frequency noise components in the time-domain response waveform are filtered out by a dynamic noise suppression method;

[0013] The filtered voltage-current parameters are compared point by point with the segmented slopes of the preset voltage-current characteristic curve, and the abnormal fluctuation segments beyond the tolerance interval are extracted and marked as characteristic spectra;

[0014] The amplitude difference and duration of adjacent fluctuation segments in the characteristic spectrum are non-linearly weighted to generate a composite deviation signal containing phase shift information;

[0015] Based on the energy distribution characteristics of the composite deviation signal, a pattern recognition classifier is used to classify the deviation types into three categories: transient overshoot, steady-state offset, and harmonic oscillation, and the corresponding error type labels are output;

[0016] Select different compensation strategies according to the error type labels; among them, transient overshoot adopts a fast suppression strategy, steady-state offset adopts a gradual adjustment strategy, and harmonic oscillation adopts a phase synchronization correction strategy;

[0017] Based on the selected compensation strategy, the reference value of the preset voltage-current characteristic curve is adjusted in real time to generate new reference parameters;

[0018] The difference between the new reference parameter and the actual detected value of the anode voltage / cathode current is used as the voltage / current deviation correction amount.

[0019] Preferably, based on the voltage / current deviation correction amount, the driving frequency of the high-frequency inverter is dynamically adjusted, and at the same time, the ripple characteristic quantity feedback from the power supply output end is collected to form a closed-loop correction factor. Specifically:

[0020] Map the voltage / current deviation correction amount to the initial adjustment amount of the driving frequency, and synchronously collect the time-domain ripple waveform at the output end of the high-frequency inverter;

[0021] Perform time-frequency ridge tracking on the time-domain ripple waveform, extract the fundamental wave deviation degree of the ripple component and the energy density distribution of the harmonic clustering interval, and construct a ripple feature vector containing amplitude-frequency coupling characteristics;

[0022] Based on the energy density distribution of the ripple feature vector, analyze the conduction loss trend of the inverter switching device, and generate a ripple suppression weight coefficient linked to the drive frequency adjustment amount;

[0023] Input the ripple suppression weight coefficient and the initial adjustment amount into a dynamic coupler, and determine the transient compensation gradient and steady-state offset compensation amount of the drive frequency according to the amplitude-phase compensation rule switched by the current working mode of the inverter;

[0024] Fuse the transient compensation gradient and the steady-state offset compensation amount according to a preset fusion ratio, and then generate a closed-loop correction factor with time-varying characteristics after phase synchronization calibration.

[0025] Preferably, couple and analyze the closed-loop correction factor with the change rate of the cathode emission current of the X-ray tube to generate a phase compensation instruction for the pulse width modulation signal, specifically:

[0026] Establish a dynamic correlation matrix between the closed-loop correction factor and the change rate of the cathode emission current, and extract the phase lag characteristics of the two in the time domain through cross-correlation analysis;

[0027] Based on the phase lag characteristics, perform piecewise smoothing processing on the waveform of the cathode current change rate to generate a current modulation coefficient with time-varying gain characteristics;

[0028] Input the current modulation coefficient and the closed-loop correction factor into a non-linear coupler, and select the corresponding coupling weight distribution strategy according to the current working mode of the X-ray tube to determine the fundamental wave phase offset of the pulse width modulation signal; wherein, the working mode includes continuous exposure mode and pulse modulation mode;

[0029] Invert the pulse duty ratio distribution within the switching period through the fundamental wave phase offset, and combine the preset dead time constraint conditions to generate a phase compensation instruction including leading edge compensation and trailing edge delay.

[0030] Preferably, select the corresponding coupling weight distribution strategy according to the current working mode of the X-ray tube to determine the fundamental wave phase offset of the pulse width modulation signal, specifically:

[0031] Establish an X-ray tube working mode recognition module, and determine whether it is currently in the continuous exposure mode or the pulse modulation mode by analyzing the duty ratio characteristics and rising edge slope of the cathode emission current;

[0032] In the continuous exposure mode, perform weighted averaging on the closed-loop correction factor and the current change rate to generate a coupling weight coefficient dominated by steady-state accuracy;

[0033] In the pulse modulation mode, generate a coupling weight coefficient dominated by fast tracking by extracting the peak interval of the current change rate and the mutation point of the closed-loop correction factor and performing dynamic matching;

[0034] Input the coupling weight coefficient into the phase shift calculator, and use the least squares method to fit the fundamental wave phase deviation trend in the continuous exposure mode, while using the peak locking method to capture the phase jump point of the dominant harmonic in the pulse modulation mode;

[0035] Finally, output the fundamental wave phase shift of the pulse width modulation signal based on the selected calculation mode; among them, the offset in the continuous exposure mode is used to correct the steady-state waveform distortion, and the offset in the pulse modulation mode is used to compensate for the phase lag during fast switching.

[0036] Preferably, reconstruct the trigger timing map of the power switch device according to the phase compensation instruction, and synchronously inject the thermodynamic parameters of the tube body for secondary duty cycle correction, specifically:

[0037] Input the phase compensation instruction into the timing reconstruction engine, separate the conduction phase reference point and the turn-off phase reference point of the power switch device by the pulse width-phase decoupling method, and generate an initial trigger timing map;

[0038] Synchronously collect the X-ray tube body temperature gradient distribution data and calculate the heat dissipation efficiency coefficient of each area of the tube body, and map the heat dissipation efficiency coefficient of each area of the tube body to the allowable maximum duty cycle attenuation gradient;

[0039] Input the initial trigger timing map and the maximum duty cycle attenuation gradient into the dynamic constraint optimizer, and use the thermal-electrical coupling algorithm to adaptively adjust the conduction phase reference point according to the temperature to generate an optimized trigger timing with thermal protection characteristics;

[0040] According to the ratio of the tube body temperature change rate to the preset thermal time constant, dynamically adjust the delay compensation amount of the turn-off phase reference point, and then generate a temperature-adaptive duty cycle correction value in combination with the duty cycle parameter of the initial trigger timing map;

[0041] Re-integrate the duty cycle correction value and the turn-off phase reference point to form a trigger timing map after secondary correction.

[0042] Preferably, generate a multi-parameter collaborative control instruction by iteratively optimizing the matching relationship between the trigger timing map and the duty cycle correction value, specifically:

[0043] Establish a dynamic correlation matrix between the trigger timing map and the duty cycle correction value, analyze the collaborative matching degree of the two in the time domain through a multi-objective optimization algorithm, and generate an initial set of collaborative control parameters;

[0044] Synchronously collect the output X-ray intensity of the X-ray tube, and extract the sensitivity coefficients of the output X-ray intensity and each control parameter;

[0045] Input the sensitivity coefficient into the adaptive weight allocator to dynamically adjust the weight ratio of the trigger timing and duty cycle in the control model according to the current working stage;

[0046] Online optimize the initial collaborative control parameter set based on historical control effect data to generate a parameter correction vector with time-varying characteristics;

[0047] Inject the parameter correction vector into the fuzzy inference engine, and combine the boundary conditions of the tube body temperature and the cathode current change rate to generate a multi-parameter collaborative control instruction.

[0048] Preferably, online optimize the initial collaborative control parameter set based on historical control effect data to generate a parameter correction vector with time-varying characteristics, specifically:

[0049] Establish a historical control effect database to store multi-dimensional historical operation data including X-ray intensity stability indicators, power supply efficiency parameters, and the thermodynamic state of the tube body;

[0050] Dynamically weight the multi-dimensional historical operation data to form a time-effect feature vector, input the time-effect feature vector into the parameter sensitivity analyzer to determine the influence weight of each collaborative control parameter on the system performance;

[0051] Generate a parameter optimization priority sequence based on the influence weight, and adopt a progressive adjustment strategy to directionally optimize the high-priority parameters;

[0052] Compare the optimization result with the real-time system feedback, dynamically update the adjustment step size and direction of the parameter correction vector, and finally generate a correction vector containing time-varying characteristic parameters.

[0053] Preferably, inject the parameter correction vector into the fuzzy inference engine, and combine the boundary conditions of the tube body temperature and the cathode current change rate to generate a multi-parameter collaborative control instruction, specifically:

[0054] Input the parameter correction vector into the fuzzy controller and convert it into a standardized fuzzy variable;

[0055] At the same time, real-time monitor the tube body temperature distribution and the cathode current change rate, and generate an allowable working interval through the boundary condition detection circuit;

[0056] Input the fuzzy variable and the allowable working interval into the fuzzy rule base, and activate the corresponding control rule set according to the current working mode of the X-ray tube;

[0057] Analyze the output control rule set to obtain the preliminary adjustment amounts of the voltage-current reference value, drive frequency, and trigger timing;

[0058] Determine the execution order according to the requirements of each preliminary adjustment amount for temperature sensitivity and response speed, and generate a multi-parameter collaborative control instruction.

[0059] In a second aspect of the present invention, a power supply control system for an X-ray generating device is disclosed. The power supply control system includes a memory and a processor. A power supply control method program for the X-ray generating device is stored in the memory. When the power supply control method program for the X-ray generating device is executed by the processor, the steps of any one of the power supply control methods for the X-ray generating device are implemented.

[0060] The present invention solves the technical defects existing in the background art and has the following beneficial effects: By establishing a multi-level collaborative control mechanism for voltage-current deviation, driving frequency, phase compensation, and thermodynamic parameters, the adaptive and precise adjustment of the power supply system is achieved, improving the stability and transient response characteristics of X-ray output, effectively suppressing power supply ripple and switching losses. At the same time, the safety and reliability of the system operation are ensured through the coupling of thermodynamic parameters, enabling the X-ray tube to obtain optimized power supply under different working conditions. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or in the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0062] Figure 1 It is the overall method flowchart of the power supply control method for this X-ray generating device;

[0063] Figure 2 It is a partial method flowchart of the power supply control method for this X-ray generating device;

[0064] Figure 3 It is the system block diagram of the power supply control system for this X-ray generating device. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0065] In order to be able to more clearly understand the above-mentioned objects, features, and advantages of the present invention, the present invention will be further described in detail below with reference to the drawings and specific embodiments. It should be noted that, without conflict, the embodiments of the present application and the features in the embodiments can be combined with each other.

[0066] Many specific details are set forth in the following description in order to fully understand the present invention. However, the present invention can also be implemented in other ways different from those described herein. Therefore, the protection scope of the present invention is not limited by the specific embodiments disclosed below.

[0067] Such as Figure 1As shown in the figure, the first aspect of the present invention discloses a power control method for an X-ray generating device, including the following steps:

[0068] S102. Real-time detect the working state parameters of the X-ray tube, and generate corresponding voltage / current deviation correction amounts according to a preset voltage-current characteristic curve;

[0069] S104. Dynamically adjust the driving frequency of the high-frequency inverter based on the voltage / current deviation correction amount, and simultaneously collect the ripple characteristic quantity feedback from the power output end to form a closed-loop correction factor;

[0070] S106. Perform coupled analysis on the closed-loop correction factor and the change rate of the cathode emission current of the X-ray tube to generate a phase compensation instruction for the pulse width modulation signal;

[0071] S108. Reconstruct the trigger timing map of the power switch device according to the phase compensation instruction, and synchronously inject the thermodynamic parameters of the tube body for secondary duty cycle correction;

[0072] S110. Generate a multi-parameter collaborative control instruction by iteratively optimizing the matching relationship between the trigger timing map and the duty cycle correction value.

[0073] It should be noted that the present invention solves the problems of poor output stability and slow response speed caused by insufficient dynamic coupling of multiple parameters in the X-ray tube power supply system. By establishing a multi-level collaborative control mechanism for voltage-current deviation, driving frequency, phase compensation, and thermodynamic parameters, the adaptive precise adjustment of the power supply system is realized, the stability and transient response characteristics of X-ray output are improved, the power supply ripple and switching losses are effectively suppressed, and at the same time, the safety and reliability of the system operation are ensured through the coupling of thermodynamic parameters, so that the X-ray tube can obtain optimized power supply under different working conditions.

[0074] Preferably, real-time detect the working state parameters of the X-ray tube, and generate corresponding voltage / current deviation correction amounts according to a preset voltage-current characteristic curve, specifically:

[0075] Parallel data acquisition is performed on the anode voltage and cathode current of the X-ray tube, and the high-frequency noise components in the time-domain response waveform are filtered out by a dynamic noise suppression method;

[0076] The filtered voltage-current parameters are compared point by point with the piecewise slopes of the preset voltage-current characteristic curve, and the abnormal fluctuation segments exceeding the tolerance interval are extracted and marked as the characteristic spectrum;

[0077] The amplitude difference and duration of adjacent fluctuation segments in the characteristic spectrum are non-linearly weighted to generate a composite deviation signal containing phase shift information;

[0078] Based on the energy distribution characteristics of the composite deviation signal, a pattern recognition classifier is used to classify the deviation types into three categories: transient overshoot, steady-state offset, and harmonic oscillation, and the corresponding error type labels are output;

[0079] Different compensation strategies are selected according to the error type labels; among them, the transient overshoot adopts a fast suppression strategy, the steady-state offset adopts a progressive adjustment strategy, and the harmonic oscillation adopts a phase synchronization correction strategy;

[0080] Based on the selected compensation strategy, the reference value of the preset voltage-current characteristic curve is adjusted in real time to generate new reference parameters (such as reducing the target voltage in the overshoot section or smoothing the current slope in the fluctuation section);

[0081] The difference between the new reference parameter and the actual detected value of the anode voltage / cathode current is used as the voltage / current deviation correction amount.

[0082] Exemplarily, a high-voltage differential probe (bandwidth ≥ 100 MHz) and a Hall effect current sensor are used to synchronously collect the anode voltage (typical value 80 kV ± 5%) and the cathode current (typical value 200 mA ± 3%). High-frequency noise is filtered by an analog active filter with a cut-off frequency of 1 / 10 of the switching frequency (for example, 20 kHz). The filtered voltage-current signal is compared in real time with the preset characteristic curve (the segmented slope tolerance is set to ±2% / μs) stored in the FPGA. When a fluctuation section that lasts for more than 10 μs and deviates from the preset value by > 5% is detected, it is marked as a characteristic spectrum. The amplitude difference (unit: V or mA) and the duration (unit: μs) of adjacent fluctuation sections in the characteristic spectrum are weighted and fused according to a weight ratio of 1:0.7 to generate a composite deviation signal with time stamp information (data update period 50 μs). Through a classifier pre-stored with 300 sets of typical working condition training samples, the composite deviation signal with a rise time < 50 μs is identified as a transient overshoot, the one with a fluctuation period of 1 - 10 ms is identified as a harmonic oscillation, and the rest are identified as steady-state offsets.

[0083] When the system detects a transient overshoot (such as an instantaneous overshoot of the anode voltage by +8 kV), the reference voltage is reduced from 80 kV to 75 kV within 100 μs through the fast suppression strategy. At this time, the voltage deviation correction amount is -5 kV; for a steady-state offset (such as the cathode current continuously being 15 mA lower), the reference current is gradually increased at a rate of 2 mA / ms, generating a current deviation correction amount of +10 mA; for a harmonic oscillation (such as a periodic fluctuation of ±3 kV), an oscillation suppression correction amount of ±1.5 kV is generated after implementing phase synchronization compensation.

[0084] In summary, the present invention solves the problem of insufficient voltage-current dynamic matching accuracy caused by load changes in X-ray tube power supply control. Through real-time noise suppression, characteristic spectrum analysis, and intelligent deviation classification, accurate identification and differential compensation of transient overshoot, steady-state offset, and harmonic oscillation are achieved, thereby improving the tracking accuracy and response speed of the working parameters of the X-ray tube. At the same time, the instability of the X-ray output intensity caused by power fluctuations is effectively suppressed, providing a high-precision deviation correction benchmark for subsequent power supply regulation.

[0085] Preferably, based on the voltage / current deviation correction amount, the driving frequency of the high-frequency inverter is dynamically adjusted, and at the same time, the ripple characteristic quantity feedback from the power supply output end is collected to form a closed-loop correction factor, as Figure 2 shown, specifically:

[0086] S202. Map the voltage / current deviation correction amount to the initial adjustment amount of the driving frequency, and synchronously collect the time-domain ripple waveform at the output end of the high-frequency inverter;

[0087] S204. Perform time-frequency ridge tracking on the time-domain ripple waveform, extract the fundamental wave offset degree of the ripple component and the energy density distribution of the harmonic clustering interval, and construct a ripple feature vector including amplitude-frequency coupling characteristics;

[0088] Among them, time-frequency ridge tracking refers to the means of performing time-frequency joint analysis on the time-domain ripple waveform, and determining the frequency drift characteristics of the ripple fundamental wave and harmonic components and the energy concentration region by extracting the continuous peak trajectory of the signal energy in the time-frequency distribution diagram.

[0089] S206. Based on the energy density distribution of the ripple feature vector, analyze the conduction loss trend of the inverter switching device, and generate a ripple suppression weight coefficient linked to the driving frequency adjustment amount;

[0090] It should be noted that based on the collected ripple feature vector (including fundamental wave offset degree and harmonic energy distribution), the system first analyzes the conduction loss situation of the switching device under the current working state. When the detected harmonic energy is higher than the preset value (indicating an increase in switching loss), a weight coefficient for reducing the driving frequency is automatically generated; otherwise, a weight coefficient for increasing the frequency adjustment amplitude is generated. This weight coefficient will change dynamically according to the ripple characteristics detected in real time and is combined with the initial frequency adjustment amount to finally form a comprehensive frequency adjustment scheme that can both suppress ripple and optimize switching loss.

[0091] S208. Input the ripple suppression weight coefficient and the initial adjustment amount into the dynamic coupler, and determine the transient compensation gradient and steady-state offset compensation amount of the driving frequency according to the amplitude-phase compensation rule switched by the current working mode (hard switching / soft switching) of the inverter;

[0092] S210. Fuse the transient compensation gradient and the steady-state offset compensation amount according to a preset fusion ratio, and then generate a closed-loop correction factor with time-varying characteristics after phase synchronization calibration.

[0093] Exemplarily, when the detected voltage deviation correction amount is +5 kV, it is converted into an initial drive frequency adjustment amount of +750 Hz according to a proportionality coefficient of 0.15 kHz / kV; at the same time, an oscilloscope with a bandwidth of 100 MHz is used to collect the ripple waveform (typical amplitude ±2 V) at the output end of the inverter. The frequency deviation amount (such as +1.2 kHz) of the ripple fundamental component (center frequency 100 kHz) and the energy ratio (such as 28%) in the second harmonic interval (190 - 210 kHz) are extracted through a fast Fourier transform processor to construct a ripple feature vector containing amplitude-frequency correlation parameters. According to the threshold condition that the harmonic energy ratio in the feature vector exceeds 25%, a drive frequency downweighting coefficient of 0.8 is generated. In the hard-switching mode, the weighting coefficient and the initial adjustment amount are coupled at a ratio of 3:7 to output a transient compensation gradient of -300 Hz / ms; in the soft-switching mode, they are coupled at a ratio of 5:5 to output a steady-state offset compensation amount of +200 Hz. Finally, the two compensation amounts are superimposed according to a 6:4 fusion ratio corresponding to the dynamic load rate (such as 70%), and a closed-loop correction factor (such as a final output adjustment amount of +420 Hz) is generated after 50-ns-level phase calibration, and the correction factor update period is 100 μs.

[0094] In summary, by establishing a dynamic coupling mechanism between voltage / current deviation and ripple characteristics, the real-time and precise adjustment of the drive frequency is achieved, solving the problems of lag in drive frequency adjustment and poor ripple suppression effect caused by load mutation in the application of high-frequency inverters in X-ray power supplies, thereby improving the dynamic response speed of the power supply system, effectively suppressing the output ripple, and at the same time reducing the conduction loss of switching devices, providing a more stable high-frequency power output for the X-ray tube.

[0095] Preferably, the closed-loop correction factor is coupled and analyzed with the change rate of the cathode emission current of the X-ray tube to generate a phase compensation instruction for the pulse width modulation signal, specifically:

[0096] Establish a dynamic correlation matrix between the closed-loop correction factor and the change rate of the cathode emission current, and extract the phase lag characteristics of the two in the time domain through cross-correlation analysis; that is, extract the phase lag characteristics of the closed-loop correction factor and the change rate of the cathode emission current in the time domain.

[0097] It should be noted that the closed-loop correction factor signal and the cathode current change rate signal are synchronously recorded, and waveform data of at least 10 complete cycles (about 100 - 500 microseconds) are stored at a sampling interval of 1 microsecond; then the two groups of signals are input into a digital processor, and the similarity between the two is calculated point by point through a sliding time window (such as 50 microseconds in width). When it is detected that the waveform of the cathode current change rate shows a similar change trend with a delay compared to the correction factor waveform (for example, the current change always peaks 15 microseconds later than the correction factor), this time difference is the phase lag feature.

[0098] Based on the phase lag feature, the waveform of the cathode current change rate is segmented and smoothed to generate a current modulation coefficient with time-varying gain characteristics;

[0099] The current modulation coefficient and the closed-loop correction factor are input into a non-linear coupler, and the corresponding coupling weight allocation strategy is selected according to the current working mode of the X-ray tube to determine the fundamental phase offset of the pulse width modulation signal; among them, the working mode includes a continuous exposure mode and a pulse modulation mode;

[0100] The pulse duty ratio distribution within the switching period is inversed through the fundamental phase offset, and combined with the preset dead-time constraint conditions, a phase compensation instruction including leading-edge compensation and trailing-edge delay is generated.

[0101] Among them, the preset dead-time constraint condition refers to the minimum time interval forcibly set in the control signal of the power switching device to ensure that the upper and lower bridge arm switching tubes do not conduct simultaneously. This time value is determined according to the turn-off characteristics of the specific switching device (usually 1 - 3 μs) and is used to prevent the power supply from being directly short-circuited and reduce the switching loss.

[0102] Preferably, the corresponding coupling weight allocation strategy is selected according to the current working mode of the X-ray tube to determine the fundamental phase offset of the pulse width modulation signal, specifically:

[0103] An X-ray tube working mode recognition module is established. By analyzing the duty ratio feature and the rising edge slope of the cathode emission current, it is determined whether the current is in the continuous exposure mode or the pulse modulation mode;

[0104] In the continuous exposure mode, the closed-loop correction factor and the current change rate are weighted and averaged to generate a coupling weight coefficient dominated by steady-state accuracy;

[0105] In the pulse modulation mode, the peak interval of the current change rate and the mutation point of the closed-loop correction factor are extracted and dynamically matched to generate a coupling weight coefficient dominated by fast tracking;

[0106] Input the coupling weight coefficient into the phase offset calculator. In the continuous exposure mode, use the least squares method to fit the fundamental wave phase deviation trend, and in the pulse modulation mode, use the peak locking method to capture the phase jump points of the dominant harmonics;

[0107] It should be noted that in the continuous exposure mode, the system will continuously collect phase deviation data for a period of time (such as 10 milliseconds), and find the overall trend line of the deviation changing with time through mathematical fitting, so as to obtain a stable phase compensation amount; while in the pulse modulation mode, the system will real-time monitor the mutation points of the current waveform. When detecting the spikes of rapid current rise or fall (usually completed within dozens of microseconds), immediately record the phase deviation value at this moment as the compensation reference. The two modes adopt different phase recognition strategies for stable working states and fast transient states respectively to ensure that the required phase compensation parameters can be accurately obtained under various working conditions.

[0108] Finally, output the fundamental wave phase offset of the pulse width modulation signal based on the selected calculation mode; among them, the offset in the continuous exposure mode is used to correct the steady-state waveform distortion, and the offset in the pulse modulation mode is used to compensate the phase lag during fast switching.

[0109] It should be noted that the system automatically selects the corresponding phase compensation strategy according to the current working mode: in the continuous exposure mode, output the steady-state phase offset obtained by smooth fitting (such as +5°) to slowly correct the waveform distortion; in the pulse modulation mode, output the transient phase jump amount captured in real time (such as -15°) to quickly compensate for the switching transient. These phase offsets will be directly converted into the time adjustment amount of the PWM signal. For example, each 1° phase offset corresponds to a pulse width change of about 28 nanoseconds, and finally generate an accurate time compensation instruction to be sent to the power switch driver.

[0110] Exemplarily, when the closed-loop correction factor is +420 Hz and the cathode current change rate reaches 50 mA / μs, the time-domain correlation matrix between the two is established by the relevant arithmetic unit of the digital signal processor, and a 15-μs phase lag of the current change relative to the correction factor is detected. A sliding average filter with a time window length of 100 μs is used to smooth the current change rate waveform segment by segment, generating a modulation coefficient (range 0.8 - 1.2) that dynamically adjusts with the current slope. In the continuous exposure mode (cathode current duty cycle > 90%), the modulation coefficient and the correction factor are coupled according to a weight ratio of 6:4, and the fundamental wave phase offset (typical value ±5°) for correcting the steady-state distortion is output; in the pulse modulation mode (rising edge slope > 80 mA / μs), they are coupled according to a weight ratio of 3:7, and the phase offset (typical value ±15°) for compensating for fast switching is output. The phase offset is converted into a duty cycle adjustment amount (resolution 0.1%) through a PWM controller, and the dead time is constrained to be ≥2 μs, and finally, a phase compensation instruction with a leading edge compensation (advanced 0.5 - 1 μs) and a trailing edge delay (delayed 0.3 - 0.8 μs) is generated. Among them, the mode recognition module automatically switches the processing strategy by comparing the current waveform characteristics (continuous mode pulse width > 10 ms, pulse mode rise time < 50 μs).

[0111] In summary, the present invention realizes precise control of the phase of the pulse width modulation signal by establishing a dynamic correlation between the closed-loop correction factor and the current change rate, and implementing a differential compensation strategy for different working modes, improves the response synchronization of the X-ray tube power supply, effectively eliminates the output power fluctuation caused by the cathode current change, and at the same time ensures the safe operation of the switching device within the dead time.

[0112] Preferably, according to the phase compensation instruction, the trigger timing map of the power switch device is reconstructed, and the thermodynamic parameters of the tube body are synchronously injected for secondary correction of the duty cycle, specifically:

[0113] The phase compensation instruction is input into the timing reconstruction engine, and the conduction phase reference point and the turn-off phase reference point of the power switch device are separated by the pulse width-phase decoupling method to generate an initial trigger timing map;

[0114] Among them, the pulse width-phase decoupling method refers to a technical means of independently analyzing and processing the pulse width parameter (duty cycle) and the phase parameter (trigger moment) in the control signal of the power switch device. By separating the reference points of the conduction and turn-off moments, independent optimization and adjustment of the two in the timing are realized.

[0115] Synchronously collect the temperature gradient distribution data of the X-ray tube body and calculate the heat dissipation efficiency coefficient of each area of the tube body, and map the heat dissipation efficiency coefficient of each area of the tube body to the maximum allowable duty cycle attenuation gradient;

[0116] Input the initial trigger timing diagram and the maximum duty cycle attenuation gradient into the dynamic constraint optimizer, and use the thermal-electric coupling algorithm to adaptively adjust the temperature of the conduction phase reference point, generating an optimized trigger timing with thermal protection characteristics.

[0117] It should be noted that the principle of the thermal-electric coupling algorithm lies in establishing a dynamic correlation model between the temperature field distribution of the X-ray tube body and the electrical parameters of the power switch device, and mapping the thermodynamic parameters (such as temperature gradient, heat dissipation efficiency) into electrical control constraint conditions (such as the maximum allowable duty cycle, conduction phase compensation amount) in real time. The core of this algorithm is to use the ratio of the temperature change rate of each region of the tube body to the preset thermal time constant to dynamically adjust the on / off timing of the switch device, so that the duty cycle and phase parameters of the trigger pulse are adaptively adjusted with temperature changes, ensuring both the stability of power output and overheat protection. Specifically, when it is detected that the local temperature rises too fast, the conduction time of the corresponding switch device in this region is automatically reduced and the trigger phase is adjusted, forming a closed-loop collaborative optimization of the thermodynamic state and electrical control parameters.

[0118] According to the ratio of the tube body temperature change rate to the preset thermal time constant, dynamically adjust the delay compensation amount of the turn-off phase reference point, and then generate a temperature-adaptive duty cycle correction value in combination with the duty cycle parameter of the initial trigger timing diagram.

[0119] Re-integrate the duty cycle correction value with the turn-off phase reference point to form a trigger timing diagram after secondary correction.

[0120] Exemplarily, when the timing reconstruction engine receives a phase compensation instruction of ±15°, it converts it into the conduction reference point (such as the original trigger moment is advanced by 1.4 μs) and the turn-off reference point (such as delayed by 0.8 μs) of the power switch device, generating an initial trigger timing (duty cycle 45%). Synchronously collect temperature data through 6 thermocouples (accuracy ±1 °C) distributed on the surface of the tube body. When it is detected that the temperature of the anode target surface area reaches 80 °C, it is calculated that the maximum duty cycle attenuation gradient in this area is 2% / °C. After inputting the initial timing and the attenuation gradient into the optimizer, when the temperature rise rate of the target surface is 5 °C / s, the conduction reference point is automatically shifted backward by 0.3 μs (the adjusted duty cycle drops to 42%). At the same time, according to the ratio 0.025 of the overall temperature change rate of the tube body (3 °C / s) to the thermal time constant (120 s), the delay compensation amount of the turn-off reference point is dynamically increased by 0.2 μs, and finally a trigger timing diagram after secondary correction is generated (the conduction reference point is shifted backward by 0.3 μs, the turn-off reference point is delayed by a total of 1.0 μs, and the actual duty cycle is adjusted to 40%).

[0121] In summary, by dynamically coupling the phase compensation instruction with the real-time tube body temperature parameter, the intelligent reconstruction of the trigger timing and the adaptive adjustment of the duty cycle are realized, solving the problem of misalignment of the trigger timing of the power switch device caused by temperature changes in the X-ray tube power supply system, thereby improving the working reliability of the power switch device, effectively preventing overheating damage of the tube body, and at the same time ensuring the stability of the X-ray output intensity, enabling the power supply system to optimize the power output efficiency in real time according to temperature changes.

[0122] Preferably, by iteratively optimizing the matching relationship between the trigger timing map and the duty cycle correction value, a multi-parameter collaborative control instruction is generated, specifically:

[0123] Establish a dynamic correlation matrix between the trigger timing map and the duty cycle correction value, analyze the collaborative matching degree of the two in the time domain through a multi-objective optimization algorithm, and generate an initial set of collaborative control parameters;

[0124] It should be noted that first, key parameters in the trigger timing map (such as the conduction phase point and the turn-off delay time) and the duty cycle correction value are recorded in real time, and the two sets of data are aligned based on the time axis; then, the cooperation effect of the two in different time periods is analyzed, such as detecting whether the trigger phase needs to be adjusted accordingly when the duty cycle increases to maintain output stability; then, according to the preset optimization objectives (such as minimum output fluctuation, highest efficiency, etc.), a set of initial control parameters that can achieve the best cooperation between the timing and the duty cycle is automatically determined (such as advancing the phase by 0.5 μs and increasing the duty cycle by 2%); finally, this set of parameters is used as the base value for subsequent optimization. The entire process is automatically updated every 100 milliseconds to ensure that the parameters always match the current working state.

[0125] Synchronously collect the output X-ray intensity of the X-ray tube, and extract the sensitivity coefficients of the output X-ray intensity and each control parameter;

[0126] Input the sensitivity coefficients into an adaptive weight allocator, and dynamically adjust the weight ratio of the trigger timing and the duty cycle in the control model according to the current working stage (start / steady state / regulation);

[0127] Based on the historical control effect data, online optimization of the initial set of collaborative control parameters is performed to generate a parameter correction vector with time-varying characteristics;

[0128] Inject the parameter correction vector into a fuzzy inference engine, and combine the boundary conditions of the tube body temperature and the cathode current change rate to generate a multi-parameter collaborative control instruction.

[0129] Exemplarily, when it is detected that the conduction phase reference point in the trigger timing diagram is +1.2 μs (corresponding to a switching frequency of 80 kHz) and the duty cycle correction value is -5%, the collaborative matching degree of the two is calculated to be 0.85 (range 0 - 1) through the correlation matrix; simultaneously, an ionization chamber sensor is used to collect the X-ray intensity (typical value 200 R / min), and it is analyzed that the sensitivity coefficient of the intensity to the trigger timing is 0.8 R / μs, and the sensitivity coefficient to the duty cycle is 1.2 R / %. In the startup phase (the first 30 seconds), the weight distributor sets the trigger timing weight to 70% and the duty cycle weight to 30%; after entering the steady state, it is adjusted to 50% each. Based on the historical data of the past 10 minutes (such as the intensity fluctuation < ±2%), a parameter correction vector (timing adjustment ±0.3 μs, duty cycle adjustment ±1.5%) is generated. When the tube body temperature exceeds 75°C or the cathode current change rate > 50 mA / ms, a restrictive control instruction (such as a maximum duty cycle of 40%) is output through fuzzy inference, and the control parameter update period is 200 ms.

[0130] In summary, by establishing a dynamic association model of trigger timing, duty cycle, and X-ray intensity, and introducing real-time sensitivity analysis and adaptive weight distribution mechanisms, intelligent collaborative control of multiple parameters is achieved, improving the stability and response speed of X-ray output intensity. At the same time, it ensures that the system can automatically maintain the optimal control state in various working stages, effectively avoiding control conflicts or performance degradation caused by traditional single-parameter regulation.

[0131] Preferably, the initial collaborative control parameter set is optimized online based on historical control effect data to generate a parameter correction vector with time-varying characteristics, specifically:

[0132] Establish a historical control effect database to store multi-dimensional historical operation data including X-ray intensity stability indicators, power supply efficiency parameters, and tube body thermodynamic states;

[0133] Perform dynamic weighting on the multi-dimensional historical operation data to form a time-dependent feature vector, and input the time-dependent feature vector into a parameter sensitivity analyzer to determine the influence weights of each collaborative control parameter (including voltage-current reference values, drive frequencies, trigger timings, etc.) on the system performance;

[0134] Generate a parameter optimization priority sequence based on the influence weights, and adopt a progressive adjustment strategy to directionally optimize high-priority parameters;

[0135] Compare the optimization results with the real-time system feedback, dynamically update the adjustment step size and direction of the parameter correction vector, and finally generate a correction vector containing time-varying characteristic parameters.

[0136] For example, the historical database stores the X-ray intensity fluctuation variance (typical value ±3%), inverter efficiency (89-92%) and tube temperature (40-85°C) data within the last 8 hours, and generates a time-dependent feature vector according to the time attenuation weight (weight of 0.7 for the last 1 hour of data). The parameter sensitivity analyzer analyzes that the trigger timing has the highest weight on the intensity stability (0.45), and the driving frequency has the second highest influence on the efficiency (0.3). The optimizer gives priority to adjusting the trigger timing parameters (step length ±0.2μs). When the real-time monitoring shows that the intensity fluctuation is reduced by 15%, the timing adjustment step in the correction vector is reduced to ±0.1μs; at the same time, the driving frequency (±0.5kHz) is adjusted in a linkage manner at a ratio of 2:1. The final output correction vector contains time-varying parameters such as timing compensation (+0.15μs) and frequency adjustment (-0.3kHz), and the weight distribution is automatically updated every 5 minutes.

[0137] Preferably, the parameter correction vector is injected into the fuzzy inference engine, and the multi-parameter coordinated control instructions are generated in combination with the boundary conditions of the tube body temperature and the cathode current change rate, specifically:

[0138] Input the parameter correction vector into the fuzzy controller and convert it into a standardized fuzzy variable;

[0139] At the same time, the temperature distribution of the tube body and the change rate of the cathode current are monitored in real time, and the allowable working range is generated through the boundary condition detection circuit;

[0140] The fuzzy variables and the allowed working range are input into the fuzzy rule base, and the corresponding control rule set is activated according to the current working mode of the X-ray tube;

[0141] Parse the output control rule set to obtain the initial adjustment of voltage-current reference value, drive frequency and trigger timing;

[0142] The execution order is determined according to the requirements of temperature sensitivity and response speed of each preliminary adjustment amount, and multi-parameter coordinated control instructions are generated.

[0143] Exemplarily, when the parameter correction vector includes a trigger timing adjustment amount of +0.15 μs and a drive frequency adjustment amount of -0.3 kHz, the fuzzy controller converts them into fuzzy variables of "medium positive timing correction" and "small negative frequency correction"; at the same time, the highest temperature of the tube body is monitored to be 82 °C (allowable upper limit 85 °C) and the cathode current change rate is 40 mA / μs (allowable upper limit 50 mA / μs), generating a safe operating range. In the continuous exposure mode, rule set A is activated, and a preliminary instruction of output voltage reference value adjustment of +0.8 kV and frequency adjustment of -0.25 kHz is output; in the pulse mode, rule set B is activated, and an instruction of output timing advanced by 0.2 μs is output. According to the temperature sensitivity ranking (timing > frequency > voltage), the timing adjustment takes effect first, the frequency adjustment is executed after 200 μs, and finally the voltage adjustment is gradually completed within 500 μs, forming a collaborative control instruction for time-sharing execution.

[0144] In this embodiment, the power supply control method may further include the following steps:

[0145] Real-time monitor the micro-variation rate of the cathode current and the fluctuation trend of the grid voltage, and generate an initial density gradient map in combination with the thermal field distribution parameters of the electron emission surface;

[0146] Input the density gradient map into the electron trajectory simulation engine, and generate electron cloud dynamic distribution parameters according to the ionization degree of the residual gas and the magnetic field compensation amount in the vacuum tube cavity;

[0147] Extract the density gradient mutation characteristics of the electron cloud based on the dynamic distribution parameters, and predict the density gradient offset amount in the next several switching cycles (such as 3 - 5 switching cycles) through the space charge relaxation time inversion algorithm;

[0148] Input the density gradient offset amount and the grid geometry parameters (including grid transparency and inter-pole curvature radius) into the feed-forward compensator to generate a non-linear feed-forward coefficient including the space charge effect suppression weight;

[0149] Perform phase-advance coupling on the feed-forward coefficient and the real-time grid bias voltage to generate a grid control bias correction instruction superimposed with the electron cloud dynamic compensation amount.

[0150] Among them, the density gradient offset amount is directly related to the generation logic of the feed-forward coefficient, and the dynamic distribution parameters output by the electron trajectory simulation engine are synchronously embedded in the calculation process of the relaxation time inversion algorithm.

[0151] Exemplarily, when it is detected that the micro-variation rate of the cathode current reaches ±5 mA / μs and the grid voltage fluctuation exceeds ±200 V, the thermal field distribution is collected through 6 infrared temperature measurement points (accuracy ±2 °C) distributed on the cathode emission surface to generate an initial density gradient map (resolution 0.1 mA / mm²). Input the map into the electron trajectory simulation engine equipped with a residual gas pressure sensor (range 10 -3 to 10-6 Pa) and the electron trajectory simulation system of the Hall magnetometer. When the measured ionization degree reaches 5×10 -4 and the axial magnetic field compensation amount is 50 mT, the dynamic distribution parameters of the electron cloud are output (update period 10 μs). Based on the density gradient mutation characteristics extracted from this parameter (such as the gradient change rate > 15% / μs), the density offset amount in the next 3 switching cycles (a total of 75 μs) is predicted (such as +12%). Combining the gate structure parameters (the grid permeability is 65% and the curvature radius is 3 mm) to generate a non-linear feedforward coefficient (range 0.8 - 1.3), and finally coupling with the real-time gate bias voltage (typical value -1 kV) to output a correction instruction (such as an advanced compensation of +150 V), and the instruction response delay < 5 μs.

[0152] In summary, this method solves the problem of the lag in gate control response caused by the space charge effect under the condition of sudden load change in the X-ray tube. By establishing a dynamic prediction model of the electron cloud density gradient and a feedforward compensation mechanism, the advanced suppression of the space charge effect is realized, the response speed and control accuracy of the gate bias voltage are improved, the X-ray output fluctuation caused by the sudden change of the electron cloud density is effectively eliminated, and at the same time, through the dynamic coupling of the residual gas ionization degree and the magnetic field compensation parameters, the adaptability of the system to the change of the complex vacuum environment is enhanced.

[0153] In this embodiment, the power supply control method may further include the following steps:

[0154] Real-time collect the junction temperature gradient, carrier concentration distribution and lattice stress wave propagation speed of the power supply control chip to obtain multi-physical field data;

[0155] Input the multi-physical field data into the quantum transport equation solver, and combine the material interface defect energy barrier topology to generate a prediction model diagram of the microscopic carrier movement trajectory;

[0156] It should be noted that the collected junction temperature, carrier concentration and lattice stress data are input into the quantum transport equation solver, and at the same time, the pre-stored material interface defect parameters (such as the atomic arrangement defect position and energy barrier value) are called; then the movement behavior of electrons in these complex environments is simulated through the principles of quantum mechanics, paying attention to the hindrance or deflection effect of the defect area on the electron movement; then according to the residence time and collision probability of electrons at different positions, the possible paths and aggregation areas of electron movement are drawn; finally, these data are integrated into a dynamic electron movement prediction diagram;

[0157] Extract the carrier concentration gradient mutation region based on the prediction model diagram of the microscopic carrier movement trajectory, and perform dynamic simulation according to the carrier concentration gradient mutation region (such as simulation software such as Sentaurus TCAD and Silvaco Atlas) to generate a carrier aggregation hot spot and a quantum tunneling probability cloud diagram;

[0158] Map the hot spot and the probability cloud map to a virtual reality interaction engine, and construct a three-dimensional dynamic warning marker layer according to device failure threshold conditions (such as Schottky barrier distortion degree, cumulative amount of impact ionization).

[0159] Through the human-computer interaction interface, the spatio-temporal coupling relationship between the carrier trajectory streamline and the warning marker is rendered in real time, and a visual dynamic cloud map covering the spatio-temporal coupling characteristics of the carrier trajectory streamline and the warning marker is obtained.

[0160] It should be noted that through the combination of multi-physical field real-time monitoring and quantum transport dynamic modeling, the visual prediction of the carrier motion trajectory and the failure risk is realized, thereby improving the spatio-temporal resolution of the power device reliability analysis, enabling designers to intuitively identify potential failure sites, and avoiding micro-failure mechanisms such as hot carrier injection and quantum tunneling breakdown in advance. At the same time, it provides accurate physical field coupling data support for device optimization.

[0161] As Figure 3 shown, the second aspect of the present invention discloses a power control system 8 of an X-ray generating device. The power control system includes a memory 60 and a processor 80. A power control method program of the X-ray generating device is stored in the memory 60. When the power control method program of the X-ray generating device is executed by the processor 80, the steps of any one of the power control methods of the X-ray generating device are implemented.

[0162] The above is only the specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention, and all of them should be covered by the protection scope of the present invention.

Claims

1. A power supply control method for an X-ray generating device, characterized in that: The following steps are involved: Real-time detection of X-ray tube working state parameters, and generation of corresponding voltage / current deviation correction according to the preset voltage-current characteristic curve; Dynamically adjusting the driving frequency of the high-frequency inverter based on the voltage / current deviation correction amount, and collecting the ripple characteristic amount fed back from the power output end to form a closed-loop correction factor; The closed-loop correction factor is coupled with the rate of change of the cathode emission current of the X-ray tube to generate a phase compensation instruction of the pulse width modulation signal; Reconstructing the trigger timing diagram of the power switch device according to the phase compensation instruction, and synchronously injecting the thermodynamic parameters of the tube body to perform secondary correction of the duty cycle; By iteratively optimizing the matching relationship between the trigger timing spectrum and the duty cycle correction value, a multi-parameter coordinated control instruction is generated.

2. A power supply control method for an X-ray generating device according to claim 1, characterized in that: Real-time detection of X-ray tube working state parameters, and generation of corresponding voltage / current deviation correction according to the preset voltage-current characteristic curve, specifically: Parallel data acquisition is performed on the anode voltage and cathode current of the X-ray tube, and the high-frequency noise components in the time domain response waveform are filtered out by a dynamic noise suppression method; The filtered voltage-current parameters are compared point by point with the segmented slope of the preset voltage-current characteristic curve, and the abnormal fluctuation segments exceeding the tolerance interval are extracted and marked as characteristic spectra; The amplitude difference and duration of adjacent fluctuation segments in the characteristic spectrum are nonlinearly weighted to generate a composite deviation signal containing phase shift information; Based on the energy distribution characteristics of the composite deviation signal, the pattern recognition classifier is used to classify the deviation types into three categories: transient overshoot, steady-state offset and harmonic oscillation, and the corresponding error type labels are output; Different compensation strategies are selected according to the error type label; among them, the transient overshoot adopts the fast suppression strategy, the steady-state offset adopts the gradual adjustment strategy, and the harmonic oscillation adopts the phase synchronization correction strategy; Based on the selected compensation strategy, the reference value of the preset voltage-current characteristic curve is adjusted in real time to generate new reference parameters; The difference between the new reference parameter and the actual detected value of the anode voltage / cathode current is used as the voltage / current deviation correction amount.

3. A power supply control method for an X-ray generating device according to claim 1, characterized in that: The driving frequency of the high-frequency inverter is dynamically adjusted based on the voltage / current deviation correction amount, and the ripple characteristic amount fed back from the power output end is collected to form a closed-loop correction factor, which is specifically: Map the voltage / current deviation correction amount to the initial adjustment amount of the driving frequency, and synchronously collect the time domain ripple waveform at the output end of the high-frequency inverter; Performing time-frequency ridgeline tracing on the time-domain ripple waveform, extracting the fundamental wave deviation of the ripple component and the energy density distribution of the harmonic clustering interval, and constructing a ripple feature vector including amplitude-frequency coupling characteristics; Based on the energy density distribution of the ripple characteristic vector, the conduction loss trend of the inverter switch device is analyzed to generate a ripple suppression weight coefficient linked to the driving frequency adjustment amount; The ripple suppression weight coefficient and the initial adjustment amount are input into the dynamic coupler, and the transient compensation gradient and the steady-state offset compensation amount of the driving frequency are determined according to the amplitude-phase compensation rule of the inverter's current working mode switching; The transient compensation gradient and the steady-state offset compensation amount are fused according to a preset fusion ratio, and then a closed-loop correction factor with time-varying characteristics is generated after phase synchronization calibration.

4. The power supply control method of an X-ray generating device according to claim 1, characterized in that: The closed-loop correction factor is coupled with the rate of change of the cathode emission current of the X-ray tube to generate a phase compensation instruction of the pulse width modulation signal, specifically: The dynamic correlation matrix between the closed-loop correction factor and the cathode emission current change rate is established, and the phase lag characteristics of the two in the time domain are extracted through cross-correlation analysis. Based on the phase lag characteristics, the cathode current change rate waveform is subjected to piecewise smoothing processing to generate a current modulation coefficient with a time-varying gain characteristic; The current modulation coefficient and the closed-loop correction factor are input into a nonlinear coupler, and a corresponding coupling weight allocation strategy is selected according to the current working mode of the X-ray tube to determine the fundamental phase offset of the pulse width modulation signal; wherein the working mode includes a continuous exposure mode and a pulse modulation mode; The pulse duty cycle distribution within the switching period is inverted by the fundamental phase offset, and a phase compensation instruction including leading edge compensation and trailing edge delay is generated in combination with the preset dead time constraint.

5. A power supply control method for an X-ray generating device according to claim 4, characterized in that: According to the current working mode of the X-ray tube, the corresponding coupling weight allocation strategy is selected to determine the fundamental phase offset of the pulse width modulation signal, which is specifically: Establish an X-ray tube working mode recognition module to determine whether it is currently in continuous exposure mode or pulse modulation mode by analyzing the duty cycle characteristics and rising edge slope of the cathode emission current; In the continuous exposure mode, the closed-loop correction factor and the current change rate are weighted averaged to generate a coupling weight coefficient dominated by steady-state accuracy. In the pulse modulation mode, the peak interval of the current change rate and the mutation point of the closed-loop correction factor are extracted and dynamically matched to generate a coupling weight coefficient dominated by fast tracking. The coupling weight coefficient is input into a phase offset calculator, and the trend of the fundamental phase deviation is fitted by the least square method in the continuous exposure mode, while the phase jump point of the dominant harmonic is captured by the peak locking method in the pulse modulation mode; Finally, the fundamental phase offset of the pulse width modulation signal is output based on the selected calculation mode; the offset in the continuous exposure mode is used to correct the steady-state waveform distortion, and the offset in the pulse modulation mode is used to compensate for the phase lag during fast switching.

6. A power supply control method for an X-ray generating device according to claim 1, characterized in that: The trigger timing diagram of the power switch device is reconstructed according to the phase compensation instruction, and the thermodynamic parameters of the tube body are synchronously injected to perform a secondary correction of the duty cycle, specifically: The phase compensation instruction is input into the timing reconstruction engine, and the on-phase reference point and the off-phase reference point of the power switch device are separated by the pulse width-phase decoupling method to generate the initial trigger timing spectrum; Synchronously collect the temperature gradient distribution data of the X-ray tube body and calculate the heat dissipation efficiency coefficient of each area of ​​the tube body, and map the heat dissipation efficiency coefficient of each area of ​​the tube body to the maximum allowable duty cycle attenuation gradient; The initial trigger timing spectrum and the maximum duty cycle attenuation gradient are input into the dynamic constraint optimizer, and the conduction phase reference point is temperature-adaptively adjusted using a thermal-electric coupling algorithm to generate an optimized trigger timing with thermal protection characteristics. According to the ratio of the tube body temperature change rate to the preset thermal time constant, the delay compensation amount of the shutdown phase reference point is dynamically adjusted, and then the duty cycle parameter of the initial trigger timing spectrum is combined to generate a temperature-adaptive duty cycle correction value; The duty cycle correction value and the turn-off phase reference point are reintegrated to form a trigger timing diagram after secondary correction.

7. A power supply control method for an X-ray generating device according to claim 1, characterized in that: By iteratively optimizing the matching relationship between the trigger timing spectrum and the duty cycle correction value, a multi-parameter coordinated control instruction is generated, specifically: A dynamic correlation matrix between the trigger timing spectrum and the duty cycle correction value is established, and the collaborative matching degree between the two in the time domain is analyzed through a multi-objective optimization algorithm to generate an initial collaborative control parameter set; Synchronously collecting the output X-ray intensity of the X-ray tube, and extracting the sensitivity coefficients of the output X-ray intensity and various control parameters; The sensitivity coefficient is input into an adaptive weight allocator to dynamically adjust the weight ratio of the trigger timing and the duty cycle in the control model according to the current working stage; Based on the historical control effect data, the initial cooperative control parameter set is optimized online to generate a parameter correction vector with time-varying characteristics; The parameter correction vector is injected into the fuzzy inference engine, and combined with the boundary conditions of tube temperature and cathode current change rate, a multi-parameter collaborative control instruction is generated.

8. A power supply control method for an X-ray generating device according to claim 7, characterized in that: Based on the historical control effect data, the initial cooperative control parameter set is optimized online to generate a parameter correction vector with time-varying characteristics, specifically: Establish a historical control effect database to store multi-dimensional historical operation data including X-ray intensity stability indicators, power efficiency parameters and tube body thermodynamic state; Dynamically weight the multi-dimensional historical operation data to form a time-sensitive feature vector, which is then input into the parameter sensitivity analyzer to determine the weight of each collaborative control parameter on system performance. Generate a parameter optimization priority sequence based on the impact weights, and use a progressive adjustment strategy to perform targeted optimization on high-priority parameters; The optimization results are compared with the real-time system feedback, and the adjustment step and direction of the parameter correction vector are dynamically updated to finally generate a correction vector containing time-varying characteristic parameters.

9. The power supply control method of an X-ray generating device according to claim 7, characterized in that: The parameter correction vector is injected into the fuzzy inference engine, and combined with the boundary conditions of tube temperature and cathode current change rate, a multi-parameter coordinated control instruction is generated, specifically: Input the parameter correction vector into the fuzzy controller and convert it into a standardized fuzzy variable; At the same time, the temperature distribution of the tube body and the change rate of the cathode current are monitored in real time, and the allowable working range is generated through the boundary condition detection circuit; The fuzzy variables and the allowed working range are input into the fuzzy rule base, and the corresponding control rule set is activated according to the current working mode of the X-ray tube; Parse the output control rule set to obtain the initial adjustment of voltage-current reference value, drive frequency and trigger timing; The execution order is determined according to the requirements of temperature sensitivity and response speed of each preliminary adjustment amount, and multi-parameter coordinated control instructions are generated.

10. A power supply control system for an X-ray generating device, characterized in that: The power control system includes a memory and a processor, wherein the memory stores a power control method program for an X-ray generating device. When the power control method program for an X-ray generating device is executed by the processor, the steps of the power control method for an X-ray generating device as described in any one of claims 1 to 9 are implemented.

Images