Ultrasonic vibrator group self-adaptive repairing system and method thereof

Abstract

The invention discloses an ultrasonic vibrator group adaptive repair system and method, and relates to the technical field of ultrasonic vibration and adaptive control, and the system comprises a fault detection and isolation module, an adjacent vibrator compensation module, an adaptive frequency matching module, a phase control compensation module and an intelligent optimization and learning module. The power, the frequency and the vibration state of an oscillator are monitored in real time, once a fault oscillator is detected, the oscillator is immediately isolated, and the power and the frequency of an adjacent oscillator are dynamically adjusted according to the fault position to realize compensation; and meanwhile, frequency matching and phase modulation are carried out by utilizing the FPGA, so that the residual normal oscillators are kept in vibration consistency. The system records and optimizes each repair data through machine learning, has the advantages of rapid fault response, high compensation precision and strong expandability, and can be widely applied to the fields of industrial nondestructive testing, medical ultrasonic imaging and the like.

Description

Technical Field

[0001] The invention relates to the technical field of ultrasonic vibration and adaptive control, in particular to an ultrasonic vibrator group adaptive repair system and method thereof. Background Art

[0002] With the widespread application of ultrasonic technology in industrial cleaning, medical imaging, non-destructive testing and other fields, how to maintain continuous and stable ultrasonic output in a multi-vibrator array and quickly adaptively repair it when an individual vibrator fails, especially in scenarios that require continuous and uninterrupted use, the failure of an individual vibrator may affect the overall effect. This has become a key problem that needs to be solved urgently in the industry.

[0003] In order to deal with the efficiency drop caused by the frequency drift or failure of the vibrator, the prior art has proposed a variety of solutions. For example, Chinese invention patent CN102105234B discloses a method of adjusting the matching between the oscillator body and the vibrator through an external connection component, so that the vibrator with the same frequency or different frequency can be replaced without internal adjustment or replacement of the oscillator body. When applying ultrasonic vibration to the cleaning liquid, this technology uses the output transformer (33) and the coil capacitor (34) to adjust the ultrasonic signal, and monitors the working status through the current detector (35), achieving a certain degree of flexibility and maintainability.

[0004] The above design achieves rapid replacement of the vibrator by externally transforming and matching the signal, but it still has certain limitations, such as the inability to perform real-time detection and adaptive repair of single-point or multi-point faults in the multi-vibrator array, the lack of a mechanism for automatic isolation and compensation of faulty vibrators, and the lack of the ability to jointly adjust the frequency and phase after a fault. These shortcomings not only affect the efficient operation of the ultrasonic system in complex environments, but also increase the frequency and cost of system downtime maintenance. Summary of the invention

[0005] The purpose of the present invention is to overcome the shortcomings of the prior art and propose an ultrasonic vibrator group adaptive repair system and method, which covers a comprehensive solution of fault detection, isolation, adjacent vibrator compensation, adaptive frequency matching, phase control compensation and intelligent learning, and can greatly improve the maintainability and adaptability while maintaining the overall performance of the system.

[0006] The object of the present invention is achieved through the following technical solutions: an ultrasonic vibrator group adaptive repair system and method thereof, a plurality of ultrasonic vibrators, the vibrators forming an array; Fault detection and isolation module, which is used to monitor the power output, frequency change and vibration status of each vibrator in real time, and isolate the faulty vibrator from the array when an abnormality is detected; The adjacent oscillator compensation module is used to select adjacent oscillators according to the fault location after the faulty oscillator is isolated, and dynamically adjust the power and frequency of the adjacent oscillators to compensate for the power loss of the faulty oscillator; The adaptive frequency matching module dynamically adjusts the driving frequency of the undamaged oscillators through the digital signal processing ability of the FPGA, so that the vibration frequencies of the entire array are consistent, thereby realizing the frequency compensation for the faulty oscillator; The phase control compensation module realizes the dynamic adjustment of the phase through the DDS module inside the FPGA, specifically including a phase accumulator, a waveform memory, and a phase adjustment unit, which are used to adjust the phase of the undamaged oscillators to make the vibration phases of the entire array consistent, thereby realizing the phase compensation for the faulty oscillator; The intelligent optimization and learning module is used to record the compensation parameters and effects of each repair, and optimize the compensation strategy through machine learning algorithms.

[0007] The fault detection and isolation module includes: a micro sensor installed on each oscillator, which is used to monitor the power output, frequency change, and vibration state of the oscillator in real time; The FPGA module is used to receive the sensor signal and judge whether the oscillator is abnormal. If it is abnormal, it cuts off the power supply or signal input of the faulty oscillator.

[0008] The adjacent oscillator compensation module increases its output power by adjusting the input voltage or driving signal frequency of the adjacent oscillator to compensate for the power loss of the faulty oscillator.

[0009] The adaptive frequency matching module is implemented through the DSP module of the FPGA, specifically including: The frequency detection unit is used to monitor the working frequency change of the undamaged oscillators in real time; The frequency adjustment unit is used to dynamically adjust the driving frequency of the undamaged oscillators to make the vibration frequencies of the entire array consistent.

[0010] The phase control compensation module is implemented through the DDS module inside the FPGA, specifically including: The phase accumulator is used to calculate the phase increment; The waveform memory is used to store the sine wave look-up table; The phase adjustment unit is used to adjust the phase of the undamaged oscillators in real time according to the oscillator state feedback by the sensor to make the vibration phases of the entire array consistent.

[0011] The intelligent optimization and learning module analyzes the historical data and current state of the array through machine learning algorithms, optimizes the compensation strategy, and records the compensation parameters and effects of each repair.

[0012] An adaptive repair method for an ultrasonic oscillator group includes the following steps: S1: Real-time monitoring, which monitors the power output, frequency change, and vibration state of each oscillator in real time; S2: Fault isolation, when an anomaly is detected, isolates the faulty oscillator from the array; S3: Proximity oscillator compensation, selects proximity oscillators according to the fault location, and dynamically adjusts the power and frequency of the proximity oscillators to compensate for the power loss of the faulty oscillator; S4: Frequency compensation, by dynamically adjusting the drive frequency of the undamaged oscillators, makes the vibration frequencies of the entire array consistent, thereby achieving frequency compensation for the faulty oscillator; S5: Phase compensation, adjusts the phase of the undamaged oscillators through phase modulation to make the vibration phases of the entire array consistent, thereby achieving phase compensation for the faulty oscillator; S6: Optimize the compensation strategy, record the compensation parameters and effects of each repair, and optimize the compensation strategy through machine learning algorithms.

[0013] The steps of dynamically adjusting the power and frequency of the proximity oscillators in step S3 include: S31: Power adjustment, by adjusting the input voltage or drive signal frequency of the proximity oscillators, increases their output power to compensate for the power loss of the faulty oscillator.

[0014] The frequency compensation step in step S4 is implemented through the digital signal processing ability of the FPGA, specifically including: S41: Frequency monitoring, monitors the working frequency change of the undamaged oscillators in real time; S42: Frequency adjustment, dynamically adjusts the drive frequency of the undamaged oscillators to make the vibration frequencies of the entire array consistent.

[0015] The phase compensation step in step S5 is implemented through the DDS module inside the FPGA, specifically including: S51: Phase calculation, uses a phase accumulator to calculate the phase increment; S52: Waveform storage, stores the sine wave lookup table through a waveform memory; S53: Phase adjustment, according to the oscillator state feedback by the sensor, adjusts the phase of the undamaged oscillators in real time to make the vibration phases of the entire array consistent; The steps of optimizing the compensation strategy in step S6 analyze the historical data and current state of the array through machine learning algorithms, record the compensation parameters and effects of each repair, to improve the adaptive ability and repair efficiency of the system.

[0016] The beneficial effects of the present invention are: 1. By deploying micro sensors on each oscillator and combining with the high-speed data processing ability of FPGA, abnormal oscillators can be quickly identified at the early stage of fault occurrence. Once a faulty oscillator is detected, it is immediately isolated through hardware-level instructions to prevent fault diffusion or impact on other normal oscillators, greatly improving the operational reliability of the system.

[0017] 2. After a certain oscillator is isolated, the system will automatically identify the physical location of the faulty oscillator and select adjacent normal oscillators for combined compensation of power and frequency. This adjacent compensation mechanism can restore and maintain the overall output level of the ultrasonic array in the shortest time, avoiding a large impact on the system performance caused by a single faulty oscillator, thus significantly enhancing the stability and adaptability of the system.

[0018] 3. The system dynamically unifies the driving frequencies of the undamaged oscillators through the DSP function inside the FPGA, combined with the frequency detection and adjustment unit, reducing the mutual interference between oscillators and improving the energy transfer efficiency of the oscillators. At the same time, phase control compensation is achieved using DDS (Direct Digital Synthesis) technology to ensure a high degree of phase consistency between oscillators, giving full play to the directivity and synthetic gain of the ultrasonic array and enhancing the energy output efficiency and working performance of the entire array.

[0019] 4. The built-in intelligent optimization and learning module of the system can record the data of each fault detection, compensation process, and repair effect, and incorporate these data into machine learning algorithms for analysis and training. During long-term operation, the system can continuously accumulate fault patterns and compensation experience, realizing faster prediction of new faults and better compensation decisions, forming a virtuous cycle of "fault - repair - learning - re-repair", and continuously improving the adaptive ability and repair efficiency of the system.

[0020] 5. For scenarios with multiple-point faults or frequent faults, the present invention adopts multi-level combined compensation strategies such as adjacent oscillator compensation, frequency matching, and phase adjustment to ensure that the overall operation of the machine will not experience a significant performance decline due to large-scale failures of local oscillators. This multi-level and hierarchical compensation and isolation method can timely block fault propagation and create a good working environment for other normal oscillators, significantly reducing the risk and loss of batch faults.

[0021] 6. Due to the modular design concept (fault detection and isolation module, adjacent oscillator compensation module, adaptive frequency matching module, phase control compensation module, intelligent optimization and learning module), this system has good scalability and compatibility and can be applied to multiple fields that require large-scale ultrasonic oscillator arrays, such as industrial non-destructive testing, medical ultrasonic imaging, ultrasonic cleaning, long-distance energy transmission, etc., providing efficient and self-adaptive ultrasonic emission and control means for these fields.

[0022] 7. The present invention can be quickly repaired after a fault occurs and maintain a high working efficiency, greatly reducing the system maintenance frequency and cost. The dynamic power and frequency compensation strategy also effectively avoids the long-term overload of individual oscillators, extends the overall service life of the oscillators, reduces equipment replacement and downtime. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 is the system architecture diagram of the present invention; Figure 2 is the working flowchart of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0024] Next, the technical solutions of the present invention will be clearly and completely described in conjunction with the embodiments. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative efforts fall within the scope of protection of the present invention.

[0025] Herein, it should be noted that the orientation concepts of "left", "right", "up", "down", "front", "back", "inside", and "outside" in the following solutions are all relative directions, and will not be listed one by one here.

[0026] Embodiment 1 As Figure 1 and Figure 2 shown, in this embodiment, we focus on demonstrating the basic architecture of the entire ultrasonic oscillator group adaptive repair system and elaborating on the functional modules of fault detection and isolation as well as adjacent oscillator compensation. These modules cooperate with each other to jointly achieve the adaptive repair and compensation of ultrasonic oscillators when a fault occurs.

[0027] This system consists of an oscillator array formed by several ultrasonic oscillators, and each oscillator can emit ultrasonic waves with a specific frequency and power. To ensure good working performance even when some oscillators fail, we have arranged multiple functional modules in the system, including: Fault detection and isolation module, adjacent oscillator compensation module, adaptive frequency matching module, phased compensation module, intelligent optimization and learning module.

[0028] In this embodiment, we focus on introducing the first three parts: fault detection and isolation, adjacent oscillator compensation, and their collaborative working mechanism with the overall system architecture. The adaptive frequency matching, phased compensation, and intelligent optimization and learning modules are also mentioned in this embodiment, but their detailed implementation will be supplemented and described in subsequent embodiments.

[0029] Overview of the working process Normal state: Each oscillator operates normally according to the preset drive signal, and the frequency, phase, and power are controlled through digital circuits such as FPGA to maintain the stable ultrasonic output of the entire array. Fault detection: The fault detection and isolation module will collect the working status of the oscillators in real time, including power, frequency offset, and vibration characteristics. Once an oscillator shows an abnormality, the system will trigger fault isolation. Compensation process: The influence of the isolated oscillator on the overall output power or radiation performance is balanced by the adjacent oscillator compensation module. Specifically, the input voltage or the drive signal frequency of the surrounding oscillators can be adjusted to increase the power output, thereby making up for the loss caused by the faulty oscillator.

[0030] Fault detection and isolation module Sensor acquisition, micro sensors: In this embodiment, we install micro sensors on each oscillator to sense the output power, frequency change, and vibration state of the oscillator. These sensors are small in size, low in power consumption, and have a high integration degree with the oscillator, enabling real-time and accurate data acquisition. FPGA judgment and isolation Data reception and analysis: The FPGA module receives the raw data from the oscillator sensors through a high-speed interface and uses built-in algorithms to determine whether there is a fault in the oscillator. Fault judgment logic: When the sensor data shows that the output power of an oscillator is significantly lower than the normal range, or the frequency drift is too large, or the vibration signal is abnormal, the system will determine that there is a potential fault in the oscillator. Automatic isolation: Once the fault is determined to be established, the FPGA module will send a disconnection instruction to the drive power supply or signal channel of the faulty oscillator, completely cutting off the power supply or signal input of the faulty oscillator, thereby "physically" isolating it from the system and no longer affecting the overall output. Fault recording and subsequent processing Data recording: The occurrence time, fault location, and relevant sensor data information of each fault will be automatically recorded by the system for subsequent analysis. Subsequent optimization: These fault data will be input into the intelligent optimization and learning module, enabling the system to more accurately predict and handle similar faults in subsequent use. Through the above design, the fault detection and isolation module of this embodiment can accurately detect and isolate the faulty oscillator in an extremely short time, laying a foundation for the subsequent compensation process.

[0031] Adjacent oscillator compensation module Adjacent oscillator identification: When an oscillator is judged to be faulty and isolated, the system needs to determine the normal oscillators adjacent to or relatively close to the faulty oscillator in order to perform power or signal compensation on them. The system selects the neighboring oscillators that are most suitable for participating in compensation according to the physical layout of the oscillators in the array (such as matrix coordinates or circular arrangement, etc.). Power and Frequency Adjustment Input voltage adjustment: In this embodiment, we mainly increase the output power of the neighboring oscillators by increasing their drive voltage or drive signal level, so as to make up for the power loss of the isolated oscillator. Fine-tuning of drive frequency: If necessary, the drive frequency of the neighboring oscillators can also be appropriately adjusted to adapt to the current overall working conditions. This kind of fine-tuning usually only needs to be completed within a small range to ensure that the neighboring oscillators can work at the optimal power point. Multi-Oscillator Cooperative Compensation In complex application scenarios, faults may not only occur in a single oscillator, but even multiple oscillators may fail simultaneously. At this time, the neighboring oscillator compensation module will check the fault situation in turn and perform "orderly and hierarchical" compensation for multiple fault points. Through the centralized management of the FPGA, the system can complete the calculation and implementation of the multi-oscillator compensation strategy in an extremely short time, and restore the overall output level of the ultrasonic array as much as possible.

[0032] Dynamic Regulation and Feedback The entire compensation process is not completed "once", but continuous dynamic regulation is carried out as the isolation of the faulty oscillator and the state changes of other oscillators. After the output of the neighboring oscillators is adjusted to the ideal level, the system will still continuously monitor the working states of each oscillator. If it is found that the neighboring oscillators show performance degradation due to long-term overloading, the strategy will also be adjusted in time to avoid the occurrence of new faults.

[0033] Cooperation with Other Modules Although in this embodiment we mainly focus on fault detection, isolation, and neighboring oscillator compensation, in fact, these functions are closely coordinated with the adaptive frequency matching module, phase control compensation module, and intelligent optimization and learning module. Briefly speaking: Adaptive frequency matching module: After neighboring oscillator compensation, in order to further improve the output consistency of the entire array, it is also necessary to uniformly fine-tune the drive frequencies of the remaining normal oscillators. Phase control compensation module: In addition to power and frequency, the phase consistency also affects the overall directivity and radiation efficiency of the array. Therefore, in subsequent embodiments, we will further show how to perform phase compensation through the DDS module inside the FPGA. Intelligent optimization and learning module: The fault detection and compensation parameters that appear in this embodiment will be collected by the system and used for the training of machine learning algorithms, providing support for subsequent fault prediction and compensation strategy optimization.

[0034] Through this embodiment, faulty oscillators can be quickly detected and isolated. Meanwhile, by using the power compensation technology of adjacent oscillators, the performance loss of the entire ultrasonic array can be minimized. This modular and scalable design has good stability and flexibility in practical applications, laying a solid foundation for more advanced frequency matching and phase compensation in subsequent embodiments.

[0035] In summary, Embodiment 1 focuses on the overall system architecture, the fault detection and isolation module, and the adjacent oscillator compensation module, demonstrating effective means to achieve real-time monitoring and rapid response using FPGA, providing basic hardware and software support for the adaptive repair of the ultrasonic oscillator group.

[0036] Embodiment 2 As Figure 1 and Figure 2 shown, in this embodiment, we further introduce in detail the adaptive frequency matching module, the phase control compensation module, and the intelligent optimization and learning module in the system on the basis of the aforementioned Embodiment 1 (which mainly realizes functions such as fault detection and isolation, and adjacent oscillator compensation). After a fault occurs, these three modules can help the remaining normal oscillators to perform deeper parameter adjustments - including the adaptive correction of frequency and phase, as well as the self-learning and strategy optimization of the system during long-term operation, so that the entire ultrasonic oscillator group can maintain high working efficiency and stability in various complex environments.

[0037] Adaptive Frequency Matching Module In Embodiment 1, when some oscillators are isolated, the power output of adjacent oscillators is increased through the adjacent oscillator compensation module to make up for the loss. However, simply increasing the power is often not enough to ensure that the entire array can maintain the optimal vibration synthesis effect. Because when the working parameters of some oscillators in the array change, the originally set working frequencies of the remaining oscillators also need to be adjusted accordingly to synchronize with the new array state, so as to avoid large frequency differences. This is the core reason for introducing the adaptive frequency matching module in this embodiment.

[0038] Module Structure Frequency Detection Unit This unit can obtain the working frequency information of the undamaged oscillators in real time, including the main vibration frequency and possible slight offsets. Similar to the sensor information in Embodiment 1, these data are also input into the FPGA for high-speed analysis. But the focus here is not to detect whether there is a fault, but to detect and record the actual vibration frequency of the currently working oscillators. Frequency Adjustment Unit After obtaining the real-time operating frequencies of each oscillator, the system needs to make unified fine-tuning of them according to the overall operating conditions. Specifically, the FPGA calculates a "target frequency" or "optimal resonance frequency" and sends corresponding adjustment instructions to each undamaged oscillator. By adjusting the timing and frequency allocation of the drive signal, all normally operating oscillators can be locked near the target frequency. Implementation of the DSP module in the FPGA At the hardware level, the Digital Signal Processing (DSP) module in the FPGA undertakes the important responsibilities of frequency analysis and real-time control. It quickly calculates and outputs adjustment parameters, enabling the drive circuit to update the drive frequency of the oscillator within an extremely short time.

[0039] Workflow Data acquisition: Collect the operating frequencies of all undamaged oscillators through the sensor network. Analysis and comparison: The DSP module in the FPGA uses specific algorithms (such as Fourier transform, band-pass filtering, or phase detection, etc.) to calculate the difference between the actual frequency of the oscillator and the ideal frequency of the array. Set the target frequency: Based on the actual measurement values and the system's default optimal resonance frequency, comprehensively considering factors such as the number of faults and compensation intensity, calculate the most suitable target frequency at present. Frequency fine-tuning: Through the control instructions issued by the DSP module, the drive signals of each undamaged oscillator are synchronously increased or decreased by a small amount of Hz or kHz level to gradually approach the target frequency range. Continuous monitoring: The adjustment is not a one-time operation. It is also necessary to continuously monitor the feedback values of each oscillator. If it is found that the adjustment is excessive or the effect is not obvious, continue to optimize. Maintain the overall frequency synchronization to avoid energy loss caused by excessive frequency differences among oscillators. Compensate for the change in frequency distribution caused by the isolation of faulty oscillators, so that the ultrasonic synthesis energy and directivity of the entire system still remain in a better state. Greatly improve the adaptive ability of the system in multi-fault scenarios and extend the operating life of the equipment.

[0040] Phase control compensation module When the output phases of the oscillators are inconsistent, the ultrasonic signals that should originally be synthesized and enhanced may be cancelled or interfered, resulting in a decrease in the final radiation efficiency.

[0041] Use of the DDS module inside the FPGA Phase accumulator, one of the commonly used core components of DDS (Direct Digital Synthesis) technology is the phase accumulator. It generates signals with different phases by continuously accumulating the phase increments. Waveform Memory It is used to store the sine wave lookup table or other required waveform data. When the phase accumulator increments to the corresponding index, the corresponding waveform value is read from the memory. Phase Adjustment Unit This is the specific device in the DDS module that performs dynamic phase compensation. According to the current phase offset of the oscillator, it real-time changes the increment or initial value of the phase accumulator, thereby adjusting the phase of the output waveform.

[0042] Phase Compensation Process Status Acquisition: Similar to frequency detection, the system also needs to monitor the phase characteristics of each oscillator through sensors, or at least the inferable phase offset. Deviation Calculation: The built-in algorithm of the FPGA compares the currently measured phase with the "array reference phase" and calculates the phase increment that each oscillator needs to compensate. Dynamic Adjustment: Once the phase correction value is calculated, the phase accumulator in the DDS module immediately makes adjustments to ensure that the output drive waveform is aligned with the ideal phase. Real-time Tracking: Phase is a quantity that evolves over time and needs to be continuously tracked. If a new phase drift occurs due to a fault or overcompensation of the oscillator, the system will also correct it again.

[0043] Enhance Array Synthesis Efficiency: On the premise of ensuring sufficient oscillator power and consistent frequency, maximize the use of phase synchronization to improve the total radiation energy. Improve Directivity: In some specific application scenarios (such as ultrasonic imaging or long-distance energy transmission), phase consistency is particularly important for beamforming and energy focusing. Dynamic Adaptation to Multiple Fault Environments: Even if multiple oscillators have different types of faults, the DDS can quickly correct the phase distribution of the remaining normal oscillators and maintain relatively stable output performance.

[0044] Intelligent Optimization and Learning Module Among the functions introduced in Example 1 and this example, many operations need to analyze and make decisions based on the historical data and immediate feedback of the system. The intelligent optimization and learning module is the core unit that provides this "learning and improvement" ability. It incorporates the fault and compensation data at each stage into the model and dynamically optimizes the detection and compensation strategies for the next fault through machine learning algorithms.

[0045] Data Source Fault Detection Record: Cooperate closely with the fault isolation module in Example 1 to obtain information such as the type, fault time, and fault environment of the faulty oscillator. Oscillator Operating Status Log: The operating data of the adjacent oscillator compensation module, as well as the adaptive frequency matching and phased compensation modules, will also be packaged and sent to the learning module. Historical Strategies and Effects: The effects of each fault occurrence and compensation strategy execution (such as the degree of energy recovery, array resonance level, etc.) will be recorded in detail.

[0046] Machine Learning Algorithms and Strategy Optimization Algorithm Type: Different machine learning algorithms can be selected according to actual needs, such as deep learning, reinforcement learning, or traditional regression models. Training Process: The system can train the existing data during the idle period or offline state to update its internal prediction and decision-making models. Real-time Inference: When a new fault appears, the intelligent optimization and learning module will immediately call the trained model to make more accurate "predictions and decisions" on the level of the fault, the number of adjacent oscillators, and the compensation method, and give suggestions for the optimal or sub-optimal compensation parameters.

[0047] Continuous Evolution After the system has the self-learning ability, each repair can accumulate experience, making it more proficient in dealing with complex and diverse fault scenarios. In the long run, this continuous evolution "closed-loop learning" method will reduce misjudgments, shorten the fault repair time, and improve the overall availability and efficiency of the ultrasonic oscillator group.

[0048] The adaptive frequency matching module and the phased compensation module are both refined tuning means based on the already confirmed faulty oscillator and the preliminary power compensation for adjacent oscillators. They complement the basic functions in the first embodiment to jointly ensure the overall stability of the system. By handing over the historical data of fault detection and isolation, adjacent compensation, adaptive frequency matching, and phased compensation to the intelligent optimization and learning module, the entire system can continuously improve the repair speed and quality in complex and changing industrial or medical environments. More Stable Energy Output: The unification of frequency and phase enables the oscillator group to achieve an ideal synthesis effect. Faster Fault Repair: Through continuous optimization by machine learning, rapid identification of faults and compensation decisions are made. Longer Equipment Life: Reasonable compensation strategies avoid overusing certain adjacent oscillators and reduce the occurrence of new faults.

[0049] Based on the first embodiment, the second embodiment comprehensively introduces the structures and processes of the adaptive frequency matching module, the phased compensation module, and the intelligent optimization and learning module. Utilize the DSP function of FPGA to ensure the rapid unification of the operating frequencies of undamaged oscillators; Implement real-time phase compensation using DDS technology to enable the entire oscillator array to operate in a highly efficient and collaborative state; Record and analyze various repair and compensation data through machine learning algorithms to continuously improve the system's ability to handle complex fault scenarios.

[0050] The comprehensive application of these technologies enables the ultrasonic oscillator group to maintain its performance indicators as much as possible when single-point or multi-point failures occur, avoiding losses caused by direct shutdown, especially in important uninterruptible usage scenarios.

[0051] Embodiment 3 As Figure 1 and Figure 2 shown, based on the previous two embodiments, this embodiment further demonstrates the adaptive repair process of the ultrasonic oscillator group at the method level. This process decomposes the entire process from fault detection to final optimization in the form of steps, making the entire repair method more systematic, intuitive, and corresponding to the previously mentioned hardware and functional modules.

[0052] Overall repair process S1: Real-time monitoring Continuously obtain the power output, frequency change, and vibration state of each oscillator through a micro sensor, and be able to detect signs of abnormal performance in the first time. The sensor is installed on the surface of each oscillator or integrated inside the oscillator, and transmits data to the FPGA or other control units in real time. Use the hardware architecture (fault detection and isolation module) in the previous embodiment for data collection and preliminary judgment. Corresponding to the fault detection and isolation module introduced in Embodiment 1, the core function of this module is fully exerted in this step, laying a data foundation for the subsequent repair process.

[0053] S2: Fault isolation Trigger condition When it is detected that the power, frequency, or vibration state of a certain oscillator exceeds the normal range, or when it is predicted by the machine learning model that it has a serious deviation, fault identification is triggered. Isolation process After the FPGA detects an abnormality, it immediately issues an instruction to turn off or cut off the drive power or input signal of the faulty oscillator. The faulty oscillator no longer works synchronously with the remaining normal oscillators, avoiding its further impact on the performance of the entire array. Record and archive Store the fault information (time, oscillator position, fault type, etc.) in the database to provide data support for subsequent intelligent optimization. This step is mainly based on the fault detection and isolation function in Embodiment 1 to ensure that potential hazards can be quickly separated after a fault occurs.

[0054] S3: Proximity oscillator compensation After the output ability of the faulty oscillator disappears, it is necessary to use the normal oscillators around it or adjacent in physical position to "fill the vacancy". Select proximity oscillators The system automatically selects the oscillator closest to the fault point and with relatively stable performance as the main compensation object according to the geometric layout of the oscillator array. Dynamically adjust power and frequency S31: Power adjustment By moderately increasing the input voltage or the amplitude of the drive signal of the proximity oscillator, its output power is increased. If necessary, the operating frequency of this oscillator can also be adjusted within a small range to seek a balance between energy output and efficiency. The FPGA or other control cores can perform closed-loop control according to the real-time state to ensure the accuracy of the compensation effect. This highly corresponds to the proximity oscillator compensation module introduced in Embodiment 1, including both power increase and certain frequency fine-tuning, so that the energy gap caused by the fault is initially filled.

[0055] S4: Frequency compensation After the proximity oscillator performs power compensation, there may be a problem of inconsistent frequencies among different oscillators in the array. To maintain the overall resonance efficiency, it is necessary to more finely unify the operating frequencies of all normal oscillators. FPGA digital signal processing ability S41: Frequency monitoring Through the internal DSP of the FPGA or other digital signal processing units, continuously monitor the real-time operating frequencies of all undamaged oscillators and evaluate their deviations from the overall ideal frequency. S42: Frequency adjustment Based on the target resonance frequency, finely adjust the drive signals of each oscillator to make them gradually converge to a consistent frequency range. The hardware and algorithm implementation of the adaptive frequency matching module have been introduced in detail in Embodiment 2. This step is the specific manifestation of this module at the method level.

[0056] S5: Phase compensation If only the frequency consistency is achieved without synchronizing the phases, interference or cancellation between oscillators may still occur. Therefore, phase compensation means are needed to enhance the effect of ultrasonic energy synthesis. Implemented through the internal DDS module of the FPGA S51: Phase calculation The system uses the phase accumulator algorithm to calculate the difference between the actual output signal of each oscillator and the array reference phase, and obtains the required phase increment. S52: Waveform storage Pre-save the sine wave table corresponding to different phases in the waveform memory for the drive signal to call at any time. S53: Phase adjustment Based on the phase difference, the control unit outputs the adjusted waveform data to the corresponding oscillator drive end to make its phase consistent with the whole. Corresponding to the operation steps of the phase control compensation module, it echoes the previous hardware design (such as DDS, phase accumulator, etc.), so that the phases of each normal oscillator can be dynamically consistent.

[0057] S6: Optimize the compensation strategy In the last step of the method flow, by recording and analyzing the data of each fault and compensation process, the overall response ability of the system is continuously iteratively improved. Machine learning algorithm analysis Input: including data such as the time when the fault occurs, the type of oscillator fault, the compensation results of adjacent oscillators, the frequency and phase adjustment effects, etc. Output: an updated or corrected compensation strategy, enabling the system to quickly give a more accurate and effective compensation decision when the next fault occurs. Application scenarios In long-term and large-scale ultrasonic oscillator array operating environments, such as industrial inspection, medical ultrasonic imaging, etc., it can greatly improve the availability of equipment and reduce maintenance costs. This part matches the intelligent optimization and learning module mentioned in Embodiment 2, showing how to continuously update the decision model at the method level, closely combining hardware data and algorithm training to form a closed loop of "fault - repair - learning - repair again".

[0058] The coordination and timing between steps Overall order S1 and S2 are the preliminary work of fault detection and isolation, which are the premise of all subsequent compensation steps. S3, S4, and S5 perform power frequency compensation, full array frequency compensation, and phase compensation in sequence, going deeper layer by layer, making the array transition from the "fault state" to the "nearly normal state". S6 performs data summary and model update after all operations are completed, providing a more mature repair plan for the next fault. For real-time or semi-real-time execution, thanks to the high parallel processing ability of FPGA, the execution of S1~S5 can be completed in a very short time, effectively coping with the sudden faults of the oscillator array. If the machine learning process in S6 involves a large model, it can be trained and upgraded during off-peak hours or low-load periods, or lightweight or incremental learning algorithms can be used to dynamically update it during runtime.

[0059] Comprehensive monitoring means combined with an efficient isolation mechanism can quickly handle faults at the initial stage, preventing the faults from further deteriorating or affecting the overall performance.

[0060] From local adjacent oscillator power compensation to frequency and phase unification across the entire array, ensure that the impact of faults on output efficiency and stability is minimized.

[0061] With the continuous accumulation of fault data and the continuous training of the machine learning model, the system will become increasingly intelligent during long-term operation, and the repair speed and accuracy for different types of faults can be improved.

[0062] Whether in industrial inspection, medical diagnosis, or other scenarios that require large-scale ultrasonic arrays, the above methods can improve the system's usability and extend the equipment's lifecycle.

[0063] Example 3 elaborates in detail the adaptive repair method for ultrasonic oscillator groups in a process-oriented manner. S1 and S2 focus on introducing the operations of real-time monitoring and fault isolation; S3 emphasizes the power and preliminary frequency compensation of adjacent oscillators; S4 and S5 further reflect the unified adjustment of frequency and phase across the entire array; S6 enables the entire repair strategy to have the ability of self-learning and continuous optimization through machine learning means. Combined with the hardware systems and functional modules described in the previous two examples, this example presents a systematic and complete fault handling and repair method, greatly improving the reliability and working efficiency of ultrasonic oscillator groups in industrial and medical fields, etc., and also providing feasible ideas and technical bases for subsequent expansion and application.

[0064] The above description is only a preferred embodiment of the present invention. It should be understood that the present invention is not limited to the form disclosed herein, should not be regarded as excluding other embodiments, but can be used in various other combinations, modifications, and environments, and can be changed within the scope of the concept described herein through the above teachings or the techniques or knowledge in related fields. And the changes and alterations made by those skilled in the art that do not depart from the spirit and scope of the present invention shall all be within the protection scope of the appended claims of the present invention.

Claims

1. An ultrasonic vibrator group adaptive repair system, characterized in that: A plurality of ultrasonic vibrators, wherein the vibrators form an array; Fault detection and isolation module, which is used to monitor the power output, frequency change and vibration status of each vibrator in real time, and isolate the faulty vibrator from the array when an abnormality is detected; The adjacent vibrator compensation module is used to select an adjacent vibrator according to the fault location after the faulty vibrator is isolated, and dynamically adjust the power and frequency of the adjacent vibrator to compensate for the power loss of the faulty vibrator; The adaptive frequency matching module dynamically adjusts the driving frequency of the undamaged vibrator through the digital signal processing capability of the FPGA to make the vibration frequency of the entire array consistent, thereby achieving frequency compensation for the faulty vibrator; The phase-controlled compensation module realizes dynamic phase adjustment through the DDS module inside the FPGA. Specifically, it includes a phase accumulator, a waveform memory, and a phase adjustment unit, which are used to adjust the phase of the undamaged vibrator to make the vibration phase of the entire array consistent, thereby realizing phase compensation for the faulty vibrator; Intelligent optimization and learning module, used to record the compensation parameters and effects of each repair, and optimize the compensation strategy through machine learning algorithms.

2. The ultrasonic vibrator group adaptive repair system according to claim 1, characterized in that: The fault detection and isolation module includes: a micro sensor installed on each vibrator for real-time monitoring of the power output, frequency change and vibration state of the vibrator; The FPGA module is used to receive the sensor signal and determine whether the oscillator is abnormal. If it is abnormal, the power supply or signal input of the faulty oscillator is cut off.

3. The ultrasonic vibrator group adaptive repair system according to claim 1, characterized in that: The adjacent vibrator compensation module increases the output power of the adjacent vibrator by adjusting the input voltage or the driving signal frequency of the adjacent vibrator to compensate for the power loss of the faulty vibrator.

4. The ultrasonic vibrator group adaptive repair system according to claim 1, characterized in that: The adaptive frequency matching module is implemented by the DSP module of FPGA, and specifically includes: A frequency detection unit, used to monitor the operating frequency changes of the undamaged vibrator in real time; The frequency adjustment unit is used to dynamically adjust the driving frequency of the undamaged vibrator to make the vibration frequency of the entire array consistent.

5. The ultrasonic vibrator group adaptive repair system according to claim 2, characterized in that: The phase control compensation module is implemented by the DDS module inside the FPGA, and specifically includes: A phase accumulator for calculating a phase increment; A waveform memory for storing a sine wave lookup table; The phase adjustment unit is used to adjust the phase of the undamaged vibrator in real time according to the vibrator state fed back by the sensor, so that the vibration phase of the entire array is consistent.

6. The ultrasonic vibrator group adaptive repair system according to claim 5, characterized in that: The intelligent optimization and learning module analyzes the historical data and current status of the array through a machine learning algorithm, optimizes the compensation strategy, and records the compensation parameters and effects of each repair.

7. An ultrasonic vibrator group adaptive repair method, characterized in that: The following steps are involved: S1: Real-time monitoring, real-time monitoring of the power output, frequency change and vibration state of each vibrator; S2: Fault isolation, when an abnormality is detected, the faulty vibrator is isolated from the array; S3: Neighboring vibrator compensation: selects neighboring vibrators according to the fault location and dynamically adjusts the power and frequency of the neighboring vibrators to compensate for the power loss of the faulty vibrator. S4: Frequency compensation, by dynamically adjusting the driving frequency of the undamaged vibrator, making the vibration frequency of the entire array consistent, thereby achieving frequency compensation for the faulty vibrator; S5: Phase compensation, adjusting the phase of the undamaged vibrator through phase modulation to make the vibration phase of the entire array consistent, thereby achieving phase compensation for the faulty vibrator; S6: Optimize the compensation strategy, record the compensation parameters and effects of each repair, and optimize the compensation strategy through machine learning algorithms.

8. The method for adaptively repairing an ultrasonic vibrator group according to claim 7, characterized in that: The step of dynamically adjusting the power and frequency of the adjacent vibrator in step S3 includes: S31: power adjustment, by adjusting the input voltage or driving signal frequency of the adjacent vibrator to increase its output power and compensate for the power loss of the faulty vibrator.

9. The method for adaptively repairing an ultrasonic vibrator group according to claim 7, characterized in that: The frequency compensation step in step S4 is implemented by the digital signal processing capability of FPGA, and specifically includes: S41: Frequency monitoring, real-time monitoring of the operating frequency changes of the undamaged oscillator; S42: Frequency adjustment, dynamically adjusting the driving frequency of the undamaged vibrators to make the vibration frequency of the entire array consistent.

10. The method for adaptively repairing an ultrasonic vibrator group according to claim 7, characterized in that: The phase compensation step in step S5 is implemented by the DDS module inside the FPGA, and specifically includes: S51: Phase calculation, using a phase accumulator to calculate a phase increment; S52: waveform storage, storing the sine wave lookup table through the waveform memory; S53: phase adjustment: according to the state of the vibrator fed back by the sensor, the phase of the undamaged vibrator is adjusted in real time to make the vibration phase of the entire array consistent; The step of optimizing the compensation strategy in step S6 analyzes the historical data and current status of the array through a machine learning algorithm, and records the compensation parameters and effects of each repair to improve the system's adaptability and repair efficiency.

Images