An ultrasonic transducer group adaptive repair system and method thereof

Abstract

The application discloses an ultrasonic transducer group adaptive repair system and method, relates to the technical field of ultrasonic vibration and adaptive control, and comprises fault detection and isolation, adjacent transducer compensation, adaptive frequency matching, phased compensation and an intelligent optimization and learning module. Through real-time monitoring of transducer power, frequency and vibration state, once a fault transducer is detected, the fault transducer is immediately isolated, and the power and frequency of adjacent transducers are dynamically adjusted according to the fault position to realize compensation; meanwhile, frequency matching and phase modulation are carried out by using an FPGA, so that the remaining normal transducers remain vibration consistency. The system records and optimizes repair data each time through machine learning, has the advantages of rapid fault response, high compensation accuracy and strong scalability, and can be widely applied to the fields of industrial nondestructive testing, medical ultrasonic imaging and the like.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of ultrasonic vibration and adaptive control, and particularly relates to an ultrasonic vibrator group adaptive repair system and a method thereof. BACKGROUND

[0002] With the wide application of ultrasonic technology in the fields of industrial cleaning, medical imaging and non-destructive testing, how to maintain continuous and stable ultrasonic output in a multi-vibrator array and quickly adaptively repair individual vibrators when they fail, especially in continuous use scenarios, the failure of individual vibrators may affect the overall effect, which has become a key problem to be solved in the industry.

[0003] To cope with the efficiency decline caused by frequency drift or failure of vibrators, the prior art has proposed various solutions, for example, Chinese invention patent CN102105234B discloses a method for adjusting the matching mode of the oscillator main body and the vibrator through an external connecting assembly, so that the same frequency or different frequency vibrators can be replaced without internal adjustment or replacement of the oscillator main body. When applying ultrasonic vibration to the cleaning liquid, the output transformer (33) and the coil-capacitor (34) are used to adjust the ultrasonic signal, and the current detector (35) is used to monitor the working state, realizing a certain degree of flexibility and maintainability.

[0004] The above design realizes the quick replacement of vibrators by transforming and matching the signal externally, but still has certain limitations, such as being unable to detect and adaptively repair single-point or multi-point failures in a multi-vibrator array in real time, lacking a mechanism for automatically isolating and compensating for failed vibrators, and lacking the ability to jointly adjust the frequency and phase after failure. These deficiencies not only affect the efficient work of the ultrasonic system in complex environments, but also increase the frequency and cost of system downtime maintenance. SUMMARY

[0005] The present application aims to overcome the deficiencies of the prior art and proposes an ultrasonic vibrator group adaptive repair system and method, which covers a comprehensive solution of fault detection, isolation, adjacent vibrator compensation, adaptive frequency matching, phased compensation and intelligent learning, and can greatly improve maintainability and adaptive ability while maintaining the overall performance of the system.

[0006] The purpose of the present application is achieved by the following technical scheme: an ultrasonic vibrator group adaptive repair system and method, a plurality of ultrasonic vibrators, the vibrators forming an array;

[0007] A fault detection and isolation module is used to monitor the power output, frequency change and vibration state of each vibrator in real time, and isolate the failed vibrator from the array when an abnormality is detected;

[0008] Adjacent vibrator compensation module, for selecting adjacent vibrators according to the fault location after the fault vibrator is isolated, and dynamically adjusting the power and frequency of the adjacent vibrators to compensate for the power loss of the fault vibrator;

[0009] Adaptive frequency matching module, dynamically adjusting the driving frequency of the undamaged vibrators through the digital signal processing capability of the FPGA, so that the vibration frequency of the entire array is consistent, thereby realizing frequency compensation for the fault vibrator;

[0010] Phase compensation module, dynamically adjusting the phase through the DDS module inside the FPGA, specifically including a phase accumulator, a waveform memory and a phase adjustment unit, for adjusting the phase of the undamaged vibrators so that the vibration phase of the entire array is consistent, thereby realizing phase compensation for the fault vibrator;

[0011] Intelligent optimization and learning module, for recording the compensation parameters and effects of each repair, and optimizing the compensation strategy through machine learning algorithm.

[0012] 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;

[0013] FPGA module for receiving sensor signals and determining whether the vibrator is abnormal, and if abnormal, cutting off the power supply or signal input of the fault vibrator.

[0014] Adjacent vibrator compensation module increases the output power of the adjacent vibrators by adjusting the input voltage or driving signal frequency, compensating for the power loss of the fault vibrator.

[0015] Adaptive frequency matching module is implemented through the DSP module of the FPGA, specifically including:

[0016] Frequency detection unit for real-time monitoring of the working frequency change of the undamaged vibrator;

[0017] Frequency adjustment unit for dynamically adjusting the driving frequency of the undamaged vibrator so that the vibration frequency of the entire array is consistent.

[0018] Phase compensation module is implemented through the DDS module inside the FPGA, specifically including:

[0019] Phase accumulator for calculating phase increment;

[0020] Waveform memory for storing a sine wave lookup table;

[0021] Phase adjustment unit for real-time adjustment of the phase of the undamaged vibrator according to the vibrator state feedback by the sensor, so that the vibration phase of the entire array is consistent.

[0022] The intelligent optimization and learning module uses machine learning algorithms to analyze the array's historical data and current status, optimize compensation strategies, and record the compensation parameters and effects of each repair.

[0023] An ultrasonic vibrator group adaptive repair method comprises the following steps:

[0024] S1: Real-time monitoring, real-time monitoring of the power output, frequency change and vibration status of each vibrator;

[0025] S2: Fault isolation: when an anomaly is detected, the faulty oscillator is isolated from the array;

[0026] S3: Neighboring vibrator compensation: selects neighboring vibrators based on the fault location and dynamically adjusts their power and frequency to compensate for the power loss of the faulty vibrator.

[0027] S4: Frequency compensation, which dynamically adjusts the driving frequency of the intact vibrator to make the vibration frequency of the entire array consistent, thereby achieving frequency compensation for the faulty vibrator;

[0028] S5: Phase compensation: adjusts the phase of the intact vibrator through phase modulation to make the vibration phase of the entire array consistent, thereby achieving phase compensation for the faulty vibrator;

[0029] S6: Optimize the compensation strategy, record the compensation parameters and effects of each repair, and optimize the compensation strategy through machine learning algorithms.

[0030] The step of dynamically adjusting the power and frequency of the adjacent vibrator in step S3 includes: S31: power adjustment, increasing the output power of the adjacent vibrator by adjusting the input voltage or driving signal frequency of the adjacent vibrator to compensate for the power loss of the faulty vibrator.

[0031] The frequency compensation step in step S4 is implemented by the digital signal processing capability of the FPGA, and specifically includes:

[0032] S41: Frequency monitoring, real-time monitoring of the operating frequency changes of undamaged oscillators;

[0033] S42: Frequency adjustment, dynamically adjusting the driving frequency of the undamaged vibrators to make the vibration frequency of the entire array consistent.

[0034] The phase compensation step in step S5 is implemented by the DDS module inside the FPGA, specifically including:

[0035] S51: Phase calculation, using the phase accumulator to calculate the phase increment;

[0036] S52: waveform storage, storing the sine wave lookup table through the waveform memory;

[0037] S53: Phase adjustment, according to the sensor feedback of the vibrator state, the phase of the undamaged vibrator is adjusted in real time, so that the vibration phase of the whole array is consistent;

[0038] The step of optimizing the compensation strategy in step S6 analyzes the historical data and current state of the array through a machine learning algorithm, records the compensation parameters and effects of each repair, and improves the adaptive ability and repair efficiency of the system.

[0039] The beneficial effects of the present application are:

[0040] 1. By deploying micro sensors on each vibrator and combining the high-speed data processing capability of FPGA, abnormal vibrators can be quickly identified in the early stage of failure. Once a faulty vibrator is detected, it is immediately isolated through hardware-level instructions to prevent the spread of failure or impact on other normal vibrators, greatly improving the operational reliability of the system.

[0041] 2. When a vibrator is isolated, the system automatically identifies the physical location of the faulty vibrator and selects adjacent normal vibrators for joint 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 significant impact on system performance caused by a single faulty vibrator, thereby significantly improving the stability and adaptability of the system.

[0042] 3. The system dynamically unifies the driving frequency of undamaged vibrators through the DSP function inside the FPGA, combined with the frequency detection and adjustment unit, to reduce mutual interference between vibrators and improve energy transfer efficiency. At the same time, the DDS (Direct Digital Synthesis) technology is used to realize phase-controlled compensation, ensuring high consistency in the phase between vibrators, fully utilizing the directivity and synthesis gain of the ultrasonic array, and improving the energy output efficiency and working performance of the entire array.

[0043] 4. The built-in intelligent optimization and learning module of the system can record data of each fault detection, compensation process and repair effect, and analyze and train these data through machine learning algorithms. In the long run, the system can continuously accumulate fault patterns and compensation experience, achieve faster prediction and more optimal compensation decisions for new faults, and form a virtuous cycle of "fault-repair-learning-repair" to continuously improve the adaptive ability and repair efficiency of the system.

[0044] 5. For multiple fault or frequent fault scenarios, the present application uses a multi-level joint compensation strategy of adjacent vibrator compensation, frequency matching, phase adjustment, etc. to ensure that the overall operation will not be significantly degraded due to large-scale failure of local vibrators. This multi-level, hierarchical compensation and isolation method can block the spread of faults and create a good working environment for other normal vibrators, significantly reducing the risk and loss of batch faults.

[0045] 6. Due to the modular design idea (fault detection and isolation module, adjacent vibrator compensation module, adaptive frequency matching module, phase control compensation module, intelligent optimization and learning module), the system has good scalability and compatibility, and can be applied to multiple fields requiring large-scale ultrasonic vibrator arrays, such as industrial non-destructive testing, medical ultrasonic imaging, ultrasonic cleaning, long-distance energy transmission and the like, providing efficient and adaptive ultrasonic wave transmission and control means for these fields.

[0046] 7. The present application can quickly repair after failure and maintain high working efficiency, greatly reducing the system maintenance frequency and cost, and the dynamic power and frequency compensation strategy also effectively avoids long-term overload of individual vibrators, prolongs the service life of the overall vibrator, and reduces equipment replacement and downtime. BRIEF DESCRIPTION OF DRAWINGS

[0047] Figure 1 is the system architecture diagram of the present application;

[0048] Figure 2 is the working flowchart of the present application. DETAILED DESCRIPTION

[0049] The technical solutions of the present application will be described in detail below in conjunction with the embodiments. Obviously, the described embodiments are only a part of the embodiments of the present application, not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0050] It is explained that the orientation concepts of "left", "right", "up", "down", "front", "back", "inside" and "outside" in the following solutions are relative directions, which will not be listed one by one.

[0051] Embodiment one

[0052] As shown in Figure 1 and Figure 2 In this embodiment, we focus on showing the basic architecture of the entire ultrasonic vibrator group adaptive repair system, and detailing the function modules of fault detection and isolation and adjacent vibrator compensation. These modules cooperate with each other to realize the adaptive repair and compensation of ultrasonic vibrators when faults occur.

[0053] The system is composed of a number of ultrasonic vibrators to form a vibrator array, each vibrator can emit ultrasonic waves of a specific frequency and power. In order to ensure good working performance when some vibrators fail, we arrange multiple function modules in the system, including:

[0054] Fault detection and isolation module, adjacent transducer compensation module, adaptive frequency matching module, phase control compensation module, intelligent optimization and learning module.

[0055] In this embodiment, we focus on the first three parts: fault detection and isolation, adjacent transducer compensation, and their working mechanism with the overall system architecture. Adaptive frequency matching, phase control compensation and intelligent optimization and learning modules are also mentioned in this embodiment, but their detailed implementation will be supplemented in subsequent embodiments.

[0056] Workflow Overview

[0057] Normal state: Each transducer works normally according to the preset driving signal, and the frequency, phase and power are controlled by digital circuits such as FPGA to maintain the stability of the entire array's ultrasonic output.

[0058] Fault detection: The fault detection and isolation module will collect the working status of the transducer in real time, including power, frequency offset and vibration characteristics. Once an abnormality occurs in a transducer, the system will trigger fault isolation.

[0059] Compensation process: The impact of the isolated transducer on the overall output power or radiation performance is balanced by the adjacent transducer compensation module. Specifically, the input voltage or driving signal frequency of the surrounding transducers can be adjusted to increase the power output, thereby making up for the loss caused by the faulty transducer.

[0060] Fault detection and isolation module

[0061] Sensor collection, micro sensor: In this embodiment, we install a micro sensor on each transducer to sense the output power, frequency change and vibration state of the transducer.

[0062] These sensors are small in size, low in power consumption and high in integration with the transducers, enabling real-time and accurate data collection.

[0063] FPGA judgment and isolation

[0064] Data reception and analysis: The FPGA module receives raw data from the sensors of each transducer through a high-speed interface and uses built-in algorithms to determine whether the transducer has a fault.

[0065] Fault judgment logic: When the sensor data shows that the output power of a transducer 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 the transducer has a potential fault.

[0066] Automatic Isolation: Once a fault is determined, the FPGA module will send a disconnection command to the driving power supply or signal channel of the faulty transducer, completely cutting off its power supply or signal input, thus "physically" isolating it from the system and no longer affecting the overall output.

[0067] Fault Recording and Subsequent Processing

[0068] Data Recording: The time of each fault occurrence, fault location, and related sensor data information are automatically recorded by the system for subsequent analysis.

[0069] Subsequent Optimization: These fault data will become the input of the intelligent optimization and learning module, enabling the system to more accurately predict and respond to similar faults in future use.

[0070] Through the above design, the fault detection and isolation module of the embodiment can accurately find and isolate faulty transducers in a very short time, laying the foundation for subsequent compensation processes.

[0071] Adjacent Transducer Compensation Module

[0072] Adjacent Transducer Identification: After a transducer is determined to be faulty and isolated, the system needs to identify the normal transducers adjacent or relatively close to the faulty transducer in order to compensate for their power or signal.

[0073] The system will select the most suitable adjacent transducers for compensation based on the physical layout of the transducers in the array, such as matrix coordinates or ring arrangement.

[0074] Power and Frequency Adjustment

[0075] Input Voltage Adjustment: In this embodiment, we mainly increase the output power of adjacent transducers by increasing their driving voltage or driving signal level, thereby making up for the power loss of the isolated transducers.

[0076] Drive Frequency Fine Tuning: If necessary, the driving frequency of adjacent transducers can also be adjusted to adapt to the current overall working conditions. This fine tuning is usually only done within a small range to ensure that adjacent transducers work at the best power point.

[0077] Multi-transducer Collaborative Compensation

[0078] In complex application scenarios, faults may occur not only on a single transducer but also on multiple transducers simultaneously. In this case, the adjacent transducer compensation module will sequentially investigate the fault conditions and perform "ordered and hierarchical" compensation for multiple fault points.

[0079] Through centralized management by the FPGA, the system can complete the calculation and implementation of multi-transducer compensation strategies in a very short time, maximizing the recovery of the overall output level of the ultrasonic array.

[0080] Dynamic regulation and feedback

[0081] The entire compensation process is not "one-time" completed, but continues to dynamically regulate as the fault transducer is isolated and the state of other transducers changes.

[0082] After the output of the adjacent transducer is adjusted to the ideal level, the system will continue to monitor the working state of each transducer. If it is found that the performance of the adjacent transducer has degraded due to long-term overloading, the strategy will be adjusted in time to avoid the generation of new faults.

[0083] Coordination with other modules

[0084] Although in this embodiment, we mainly focus on fault detection, isolation and adjacent transducer compensation, in fact, these functions are closely coordinated with the adaptive frequency matching module, the phased compensation module and the intelligent optimization and learning module. Briefly:

[0085] Adaptive frequency matching module: After the compensation of the adjacent transducer, further uniform fine-tuning of the driving frequency of the remaining normal transducers is needed to further improve the output consistency of the entire array.

[0086] Phased compensation module: In addition to power and frequency, the consistency of phase 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.

[0087] Intelligent optimization and learning module: The fault detection and compensation parameters that appear in this embodiment are collected by the system and used for training of the machine learning algorithm, providing support for subsequent fault prediction and compensation strategy optimization.

[0088] Through this embodiment, the fault transducer can be quickly detected and isolated, and at the same time, the power compensation technology of the adjacent transducer is used to minimize the performance loss of the entire ultrasonic array. The modular and scalable design has good stability and flexibility in practical application, and lays a solid foundation for more advanced frequency matching and phase compensation in subsequent embodiments.

[0089] In summary, embodiment one focuses on the overall architecture of the system, the fault detection and isolation module, and the adjacent transducer compensation module, demonstrating an effective means of realizing real-time monitoring and rapid response using FPGA, providing basic hardware and software support for adaptive repair of ultrasonic transducer groups.

[0090] Embodiment two

[0091] As Figure 1 and Figure 2As shown, in this embodiment, we further introduce the adaptive frequency matching module, the phase control compensation module and the intelligent optimization and learning module based on the aforementioned embodiment one (which mainly realizes the functions of fault detection and isolation, adjacent vibrator compensation, etc.). These three modules can help the remaining normal vibrators to make deeper parameter adjustments after a fault occurs, including adaptive correction of frequency and phase, as well as self-learning and strategy optimization of the system in long-term operation, so that the entire ultrasonic vibrator group can maintain high working efficiency and stability in various complex environments.

[0092] Adaptive frequency matching module

[0093] In embodiment one, when some vibrators are isolated, the power output of adjacent vibrators will be increased by the adjacent vibrator compensation module to make up for the loss. However, simply relying on power enhancement is often not enough to ensure that the entire array can maintain the optimal vibration synthesis effect. Because when the working parameters of some vibrators in the array change, the working frequency set by the remaining vibrators also needs to be adjusted accordingly to synchronize with the new array state, so as to avoid large frequency differences. This is the core reason why we introduce the adaptive frequency matching module in this embodiment.

[0094] Module structure

[0095] Frequency detection unit

[0096] This unit can obtain the working frequency information of undamaged vibrators in real time, including the main vibration frequency and possible slight deviation. Similar to the sensor information in embodiment one, these data are also input to the FPGA for high-speed analysis. However, the focus here is not to detect whether there is a fault, but to detect and record the actual vibration frequency of the current working vibrator.

[0097] Frequency adjustment unit

[0098] After obtaining the real-time working frequency of each vibrator, the system needs to fine-tune them according to the overall operation. Specifically, the FPGA will calculate a "target frequency" or "optimal resonance frequency" and send corresponding adjustment instructions to each undamaged vibrator. By adjusting the timing and frequency allocation of the driving signal, all normal working vibrators can be locked around the target frequency.

[0099] FPGA DSP module implementation

[0100] At the hardware level, the digital signal processing (DSP) module inside the FPGA undertakes the important responsibility of frequency analysis and real-time control. It quickly calculates and outputs adjustment parameters, allowing the driving circuit to update the vibrator driving frequency in a very short time.

[0101] Workflow

[0102] Data collection: The operating frequencies of all undamaged oscillators are collected through the sensor network.

[0103] Analysis and comparison: The DSP module within the FPGA uses specific algorithms (such as Fourier transform, bandpass filtering, or phase detection) to calculate the difference between the actual frequency of the oscillator and the ideal frequency of the array.

[0104] Set target frequency: Calculate the most appropriate target frequency based on the actual measurement value and the system's default optimal resonance frequency, taking into account factors such as the number of faults and compensation strength.

[0105] Frequency fine-tuning: Through the control instructions issued by the DSP module, the drive signal of each undamaged vibrator is synchronously increased or decreased by a small amount of Hz or kHz level to gradually approach the target frequency range.

[0106] Continuous monitoring: Adjustment is not a one-time process. 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.

[0107] Maintain overall frequency synchronization to avoid energy loss caused by large frequency differences between oscillators.

[0108] Compensate for the change in frequency distribution caused by the isolation of the faulty vibrator, so that the ultrasonic synthesis energy and directivity of the entire system remain in an optimal state.

[0109] It greatly improves the system's adaptability in multiple fault scenarios and extends the equipment's operating life.

[0110] Phase control compensation module

[0111] When the output phases of the vibrators are inconsistent, the ultrasonic signals that should be synthesized and enhanced may be offset or interfered, resulting in a decrease in the final radiation efficiency.

[0112] Use of DDS module inside FPGA

[0113] A phase accumulator is a core component commonly used in DDS (Direct Digital Synthesis) technology. It generates signals of different phases by continuously accumulating phase increments.

[0114] Waveform Memory

[0115] Used to store sine wave lookup tables or other required waveform data. When the phase accumulator increments to the corresponding index, the corresponding waveform value is read from the memory.

[0116] Phase adjustment unit

[0117] This is the specific device that performs dynamic phase compensation in the DDS module. According to the current phase offset of the oscillator, it changes the increment or initial value of the phase accumulator in real time, thereby adjusting the phase of the output waveform.

[0118] Phase compensation process

[0119] State acquisition: Similar to frequency detection, the system also needs to monitor the phase characteristics of each oscillator through sensors, or at least the inferred phase offset.

[0120] Deviation calculation: The FPGA's built-in algorithm compares the currently measured phase with the "array reference phase" and calculates the phase increment that needs to be compensated for each oscillator.

[0121] 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.

[0122] Real-time tracking: Phase is a variable that evolves over time and requires continuous tracking. If the oscillator experiences new phase drift due to failure or overcompensation, the system will re-correct it.

[0123] Enhance array synthesis efficiency: Under the premise of ensuring sufficient oscillator power and consistent frequency, maximize the use of phase synchronization to increase the total radiation energy.

[0124] Improve directivity: In certain application scenarios (such as ultrasound imaging or long-distance energy transmission), phase consistency is particularly important for beamforming and energy focusing.

[0125] Dynamically adapt to multiple fault environments: Even if different types of faults occur in multiple oscillators, DDS can be used to quickly correct the phase distribution of the remaining normal oscillators to maintain relatively stable output performance.

[0126] Intelligent optimization and learning module

[0127] In the various functions described in Example 1 and this example, many operations require analysis and decision-making based on the system's historical data and immediate feedback. The intelligent optimization and learning module is the core unit that provides this "learning and improvement" capability. It incorporates fault and compensation data from each stage into the model and, using machine learning algorithms, dynamically optimizes the detection and compensation strategy for the next fault.

[0128] Data Source

[0129] Fault detection record: works closely with the fault isolation module in the first embodiment to obtain information such as the type of the faulty vibrator, fault time, and fault environment.

[0130] Transducer Operational Status Log: The operational data of the neighboring transducer compensation module and the adaptive frequency matching and phase control compensation module are also packaged and sent to the learning module.

[0131] Historical Strategy and Effect: The effect of each fault occurrence and compensation strategy execution (such as energy recovery degree, array resonance level, etc.) is recorded in detail.

[0132] Machine Learning Algorithm and Strategy Optimization

[0133] Algorithm Type: Different machine learning algorithms can be selected according to actual needs, such as deep learning, reinforcement learning, or traditional regression models.

[0134] Training Process: The system can train existing data during idle periods or offline, updating its internal prediction and decision-making models.

[0135] Real-time Inference: When a new fault occurs, the intelligent optimization and learning module immediately invokes the trained model to make more accurate "predictions and decisions" on the fault level, number of neighboring transducers, and compensation method, and provides optimal or suboptimal compensation parameter suggestions.

[0136] Continuous Evolution

[0137] After the system has self-learning ability, each repair can accumulate experience, making it more skilled in dealing with complex and diverse fault scenarios.

[0138] In the long run, this continuous evolution "closed-loop learning" approach will reduce misjudgments, shorten fault repair time, and improve the overall availability and efficiency of the ultrasonic transducer group.

[0139] The adaptive frequency matching module and the phase control compensation module are further fine-tuning means based on the confirmed fault transducers and preliminary power compensation of neighboring transducers. They complement the basic functions in Example One, together ensuring the overall stability of the system.

[0140] By giving the intelligent optimization and learning module the historical data of fault detection and isolation, neighboring compensation, adaptive frequency matching, and phase control compensation, the entire system can continuously improve repair speed and quality in complex and variable industrial or medical environments.

[0141] More stable energy output: The unity of frequency and phase allows the transducer group to achieve ideal synthesis.

[0142] Faster fault repair: Through machine learning for continuous optimization, faults are quickly identified and compensation decisions are made.

[0143] Longer equipment life: Reasonable compensation strategies avoid overusing certain neighboring transducers, reducing the occurrence of new faults.

[0144] Example Two is based on Example One, and comprehensively introduces the structure and process of the adaptive frequency matching module, the phased compensation module, and the intelligent optimization and learning module.

[0145] The DSP function of FPGA is used to ensure the rapid unification of the working frequency of undamaged oscillators.

[0146] The DDS technology is used to realize real-time phase compensation, so that the entire oscillator array can operate in a state of high efficiency and synergy.

[0147] Through machine learning algorithms, various repair and compensation data are recorded and analyzed to continuously improve the system's ability to cope with complex fault scenarios.

[0148] The comprehensive use of these technologies allows the ultrasonic oscillator group to maintain its performance indicators as much as possible when single-point or multi-point faults occur, avoiding losses caused by direct shutdown, especially in important and uninterrupted use cases.

[0149] Example Three

[0150] As shown in Figure 1 and Figure 2 , this embodiment further demonstrates the adaptive repair process of the ultrasonic oscillator group from a methodological perspective based on the previous two embodiments. This process breaks down the entire process from fault detection to final optimization in a step-by-step manner, making the entire repair method more systematic and intuitive, and corresponding to the hardware and functional modules mentioned earlier.

[0151] Overall repair process

[0152] S1: Real-time monitoring

[0153] Through micro sensors, the power output, frequency change, and vibration state of each oscillator are continuously obtained, allowing for the detection of performance abnormalities in the first instance.

[0154] Sensors are installed on the surface of each oscillator or integrated into the oscillator, and real-time data transmission to FPGA or other control units is achieved.

[0155] Data collection and preliminary judgment are performed using the hardware architecture (fault detection and isolation module) in the aforementioned embodiments.

[0156] Corresponding to the fault detection and isolation module introduced in Example One, the core functions of this module are fully utilized in this step, laying a data foundation for the subsequent repair process.

[0157] S2: Fault isolation

[0158] Triggering conditions

[0159] When the power, frequency or vibration state of a certain vibrator is monitored to be out of the normal range, or it is predicted by a machine learning model that it has deviated seriously, the fault recognition is triggered.

[0160] Isolation process

[0161] After the FPGA detects an anomaly, it immediately issues an instruction to shut down or cut off the driving power or input signal of the faulty vibrator.

[0162] The faulty vibrator no longer works in synchronization with the remaining normal vibrators, avoiding further impact on the performance of the entire array.

[0163] Recording and archiving

[0164] The fault information (time, vibrator location, fault type, etc.) is stored in the database to provide data support for subsequent intelligent optimization.

[0165] This step is mainly based on the fault detection and isolation function in Example One, to ensure that hidden dangers can be quickly isolated after a fault occurs.

[0166] S3: Compensation by adjacent vibrators

[0167] After the output capacity of the faulty vibrator disappears, it needs to be "filled in" by the normal vibrators in its vicinity or physically adjacent to it.

[0168] Selecting adjacent vibrators

[0169] The system automatically selects the vibrators closest to the fault point and relatively stable in performance as the main compensation objects based on the geometric layout of the vibrator array.

[0170] Dynamic adjustment of power and frequency

[0171] S31: Power adjustment

[0172] By moderately increasing the input voltage or driving signal amplitude of the adjacent vibrators, their output power is increased.

[0173] If necessary, the working frequency of the vibrator can also be adjusted within a small range to seek a balance between energy output and efficiency.

[0174] The FPGA or other control core can perform closed-loop control based on real-time state to ensure the accuracy of compensation effect.

[0175] This corresponds to the adjacent vibrator compensation module introduced in Example One, both including power enhancement and certain frequency fine-tuning, so that the energy gap caused by the fault is initially compensated.

[0176] S4: Frequency compensation

[0177] When the adjacent transducers are power compensated, the problem of inconsistent frequencies between different transducers may occur in the array. In order to maintain the overall resonance efficiency, the working frequency of all normal transducers needs to be more finely unified.

[0178] FPGA digital signal processing capability

[0179] S41: Frequency monitoring

[0180] Through the internal DSP or other digital signal processing units of FPGA, the real-time working frequency of all undamaged transducers is continuously monitored, and the deviation of them from the overall ideal frequency is evaluated.

[0181] S42: Frequency adjustment

[0182] Based on the target resonance frequency, the driving signal of each transducer is fine-tuned to gradually converge to a consistent frequency range.

[0183] The hardware and algorithm implementation of the adaptive frequency matching module have been described in detail in Embodiment Two, and this step is the specific embodiment of the module at the method level.

[0184] S5: Phase compensation

[0185] If only the consistency of the frequency is achieved without synchronizing the phase, it may still lead to interference or cancellation between transducers. Therefore, phase compensation means are needed to enhance the effect of ultrasonic energy synthesis.

[0186] Through the internal DDS module of FPGA

[0187] S51: Phase calculation

[0188] The system uses the phase accumulator algorithm to calculate the difference between the actual output signal of each transducer and the array reference phase, and obtains the required phase increment.

[0189] S52: Waveform storage

[0190] The corresponding sine wave table under different phases is pre-stored in the waveform storage for the driving signal to call at any time.

[0191] S53: Phase adjustment

[0192] The control unit outputs the adjusted waveform data to the corresponding transducer driving end based on the phase difference, so that the phase of each normal transducer is consistent with the whole.

[0193] The operation steps corresponding to the phase compensation module are in harmony with the previous hardware design (DDS, phase accumulator, etc.), so that the phase of each normal transducer can be dynamically consistent.

[0194] S6: Optimization compensation strategy

[0195] In the last step of the method process, by recording and analyzing the data of each failure and compensation process, the overall response capability of the system is continuously improved through iteration.

[0196] Machine Learning Algorithm Analysis

[0197] Input: including data such as the time when the fault occurred, the type of vibrator fault, the compensation results of adjacent vibrators, and the frequency and phase adjustment effects.

[0198] Output: Updated or revised compensation strategy, enabling the system to quickly make more accurate and effective compensation decisions when the next fault occurs.

[0199] Application Scenario

[0200] Long-term, large-scale ultrasonic vibrator array operation environments, such as industrial inspection and medical ultrasound imaging, can significantly improve equipment availability and reduce maintenance costs.

[0201] This part matches the intelligent optimization and learning module mentioned in Example 2, showing how to continuously update the decision model from a methodological level, closely integrating hardware data with algorithm training to form a closed loop of "fault-repair-learning-repair again".

[0202] Coordination and timing between steps

[0203] Overall order

[0204] S1 and S2 are the preliminary work for fault detection and isolation, and are the prerequisite for all subsequent compensation steps.

[0205] S3, S4, and S5 perform power frequency compensation, full array frequency compensation, and phase compensation in sequence, going deeper and deeper, allowing the array to transition from a "fault state" to a "near-normal state."

[0206] After all operations are completed, S6 will summarize the data and update the model to provide a more mature repair solution for the next failure.

[0207] Thanks to the high parallel processing capability of FPGA, the execution of S1 to S5 can be completed in a very short time, effectively coping with sudden failures of the vibrator array.

[0208] If the model in the machine learning process in S6 is large, you can choose to train and upgrade it during offline or low-load periods, or use lightweight or incremental learning algorithms to dynamically update it during runtime.

[0209] Comprehensive monitoring methods coupled with efficient isolation mechanisms enable rapid resolution of faults in their early stages, preventing them from further deteriorating or impacting overall performance.

[0210] From local adjacent transducer power compensation to frequency and phase unification across the entire array, the impact of faults on output efficiency and stability is minimized.

[0211] As fault data accumulates and machine learning models continue to train, the system becomes increasingly intelligent over time, improving both the speed and accuracy of repairs for different types of faults.

[0212] Whether in industrial testing, medical diagnosis, or other applications requiring large-scale ultrasound arrays, the above methods can improve system usability and extend device lifetimes.

[0213] Embodiment three details the ultrasound transducer group adaptive repair method in a streamlined manner.

[0214] S1 and S2 focus on real-time monitoring and fault isolation operations;

[0215] S3 emphasizes power and preliminary frequency compensation of adjacent transducers;

[0216] S4 and S5 further embody the unification of frequency and phase across the entire array;

[0217] S6 enables the entire repair strategy to have self-learning and continuous optimization capabilities through machine learning methods.

[0218] In combination with the hardware systems and functional modules described in the previous two embodiments, this embodiment presents a systematic and complete fault handling and repair method, significantly improving the reliability and efficiency of ultrasound transducer groups in industrial and medical fields, and providing a feasible idea and technical foundation for subsequent expansion and application.

[0219] The above is only the preferred embodiment of the present application, it should be understood that the present application is not limited to the form disclosed herein, should not be seen as excluding other embodiments, and can be used in various other combinations, modifications and environments, and can be modified within the scope of the concepts described herein, by the above teachings or related art or knowledge. The modifications and changes made by those skilled in the art without departing from the spirit and scope of the present application shall be within the scope of protection of the appended claims of the present application.

Claims

1. An ultrasonic vibrator group adaptive repair system, characterized in that: A plurality of ultrasonic vibrators, wherein the vibrators form an array; The fault detection and isolation module is used to monitor the power output, frequency variation, and vibration status of each transducer in real time, and isolate the faulty transducer from the array when an anomaly is detected; The adjacent vibrator compensation module is used to select adjacent vibrators according to the fault location after the faulty vibrator is isolated, and dynamically adjust the power and frequency of the adjacent vibrators to compensate for the power loss of the faulty vibrator; The adaptive frequency matching module uses the FPGA's digital signal processing capabilities to dynamically adjust the driving frequency of undamaged oscillators to ensure consistent vibration frequencies across the entire array, thereby achieving frequency compensation for faulty oscillators. The phase-controlled compensation module uses the DDS module within the FPGA to dynamically adjust the phase. Specifically, it includes a phase accumulator, waveform memory, and phase adjustment unit. It is used to adjust the phase of the undamaged vibrator to align the vibration phase of the entire array, thereby achieving phase compensation for the faulty vibrator. 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.

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 vibrator's power output, frequency change and vibration status; The FPGA module is used to receive sensor signals and determine whether the oscillator is abnormal. If 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 driving signal frequency of the adjacent vibrator, thereby compensating 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, specifically including: Frequency detection unit, used to monitor the operating frequency changes of undamaged oscillators in real time; The frequency adjustment unit is used to dynamically adjust the driving frequency of the undamaged vibrators 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-controlled compensation module is implemented by the DDS module inside the FPGA, specifically including: A phase accumulator for calculating phase increments; 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 status 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 machine learning algorithms, optimizes the compensation strategy, and records the compensation parameters and effects of each repair.

7. A method for adaptively repairing an ultrasonic vibrator group, characterized by: The following steps are involved: S1: Real-time monitoring, real-time monitoring of the power output, frequency change and vibration status of each vibrator; S2: Fault isolation: when an anomaly is detected, the faulty oscillator is isolated from the array; S3: Neighboring vibrator compensation: selects neighboring vibrators based on the fault location and dynamically adjusts their power and frequency to compensate for the power loss of the faulty vibrator. S4: Frequency compensation, which dynamically adjusts the driving frequency of the intact vibrator to make the vibration frequency of the entire array consistent, thereby achieving frequency compensation for the faulty vibrator; S5: Phase compensation: adjusts the phase of the intact 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, increasing the output power of the adjacent vibrator by adjusting the input voltage or driving signal frequency of the adjacent vibrator to 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 the FPGA, and specifically includes: S41: Frequency monitoring, real-time monitoring of the operating frequency changes of undamaged oscillators; 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, specifically including: S51: Phase calculation, using the phase accumulator to calculate the phase increment; S52: waveform storage, storing the sine wave lookup table through the waveform memory; S53: Phase adjustment: According to the vibrator status 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