Closed-loop self-adjusting method for online defect identification and real-time feedback in additive manufacturing process of titanium alloy component based on intelligent perception

Abstract

The invention discloses a closed-loop self-adjusting method for online defect identification and real-time feedback in the additive manufacturing process of a titanium alloy component based on intelligent sensing, belongs to the field of laser additive manufacturing, and solves the key problems of insufficient defect identification capability, quality control lag and the like in the current additive manufacturing process. The method is based on a multi-mode sensor architecture, integrates an infrared thermal imager, a molten pool camera and a spectrum sensor, and synchronously monitors a molten pool thermal field, radiation characteristics and surface topography. Traditional image processing and signal analysis are adopted to extract defect features such as air holes and cracks, and calibration is implemented in combination with off-line detection such as industrial CT. The recognition result is analyzed in real time through the central control platform, causes are automatically diagnosed based on defect types and distribution characteristics, parameters such as laser power and scanning paths are dynamically adjusted by matching a preset control strategy, and defect suppression and online repair are achieved. And a closed-loop self-adaptive dynamic adjustment mechanism improves the forming quality and reliability. The method is suitable for aerospace and other scenes with extremely high requirements for the quality of large complex titanium alloy components.

Description

Technical Field

[0001] The present invention belongs to the field of laser additive manufacturing, and in particular relates to a closed-loop self-adjustment method for online defect identification and real-time feedback in a multi-beam high-power laser additive manufacturing process. Background Art

[0002] During the additive manufacturing process, the intense, non-steady-state thermal cycling induced by high-energy laser beams causes the substrate surface to undergo multiple remelting and solidification cycles in a short period of time, resulting in non-equilibrium cyclic solid-state phase transitions and a high risk of various defects. For example, when the raw material powder or wire contains gas, the solidification rate of the molten pool during manufacturing is faster than the gas escape rate, and pores are likely to appear in the component. When the laser power is insufficient or the scanning speed is too fast, the raw material is not completely melted, resulting in unfused defects. When the laser's interaction time with the raw material is short, the melting, solidification, and cooling rates of the molten pool and its surrounding area are rapid, resulting in greater thermal stress and the initiation of cracks. The presence of defects can significantly affect the mechanical properties of the component, seriously affecting its service life and reliability, and limiting its further application.

[0003] Current detection methods primarily rely on offline techniques, such as industrial CT and metallographic analysis. While these technologies offer high accuracy and precision, they are expensive and result in delayed, incomplete results that fail to reflect the forming process in a timely manner. This leads to unavoidable waste and restricts process optimization efficiency. Therefore, a control system for online defect identification and real-time feedback during laser additive manufacturing is urgently needed. This system can detect components in real time during the manufacturing process and adjust process parameters in a timely manner, thereby improving the forming quality of laser additive manufacturing components. Summary of the Invention

[0004] The purpose of the present invention is to provide a closed-loop self-adjustment method for online identification and real-time feedback of defects in the multi-beam high-power laser additive manufacturing process of titanium alloy components based on intelligent perception. It aims to solve the problems of insufficient extraction of defect feature information and difficulty in online evaluation and feedback control of printing quality in the current laser additive manufacturing process, and to improve the forming quality and reliability of complex titanium alloy components (such as large suspensions, aerospace warhead casings, etc.).

[0005] To achieve the above object, the present invention provides the following technical solutions:

[0006] The present invention proposes a closed-loop self-adjustment method for online defect identification and real-time feedback in a multi-beam high-power laser additive manufacturing process of titanium alloy components based on intelligent perception, the method comprising the following steps:

[0007] Step S1: using a multi-source sensing system to synchronously collect the thermal field, light radiation, and morphological characteristics of the molten pool during the laser additive manufacturing process to obtain the original physical data of the deposited layer;

[0008] Step S2: performing defect identification on the original physical data of the deposited layer to extract the type, size, location and distribution information of the defects in the deposited layer;

[0009] Step S3: Perform offline testing on the sample to obtain the true metallurgical defect information inside the deposited layer of the sample, compare it with the identification result of step S2, and correct the defect information;

[0010] Step S4: The central computing and control platform performs cause diagnosis based on the corrected defect information and matches a process control strategy;

[0011] Step S5: Adjusting process parameters based on the control strategy to complete online repair or suppression of typical defects;

[0012] Step S6: Compare the control result with the offline detection data of the final workpiece, and realize adaptive closed-loop dynamic adjustment by optimizing the control strategy.

[0013] Furthermore, the multi-source perception system includes an infrared thermal imager, an industrial camera, and a spectral sensor;

[0014] The infrared thermal imager is used to monitor the temperature distribution changes in the molten pool and its surrounding areas;

[0015] The industrial camera is used to obtain an image of the surface of the sediment layer;

[0016] The spectrum sensor is used to sense the light radiation characteristics generated when the laser interacts with the material.

[0017] Furthermore, the above-mentioned defect identification adopts grayscale threshold method, edge detection method, image contour analysis or spectrum energy analysis.

[0018] Furthermore, the above-mentioned offline detection includes industrial CT scanning, metallographic structure analysis, metal fracture observation or scanning electron microscope imaging.

[0019] Furthermore, the above-mentioned defect types include cracks, pores, metallurgical inclusions, edge collapse, uneven thickness of the formed layer or abnormal surface undulations.

[0020] Furthermore, the above-mentioned central computing and control platform is embedded with a defect type and process causal relationship rule library, a process parameter historical data storage module, a defect statistical analysis module and a control instruction generation module.

[0021] Furthermore, the above-mentioned process control strategies include but are not limited to one or more of the following: dynamic adjustment of laser power, local switching of wire feeding modules, path redivision, and selective remelting.

[0022] The present invention also provides an intelligent perception system based on the above-mentioned closed-loop self-adjustment method for online defect identification and real-time feedback in the multi-beam high-power laser additive manufacturing process of titanium alloy components based on intelligent perception, the system comprising:

[0023] Multimodal sensor module for collecting infrared thermal images, visible light images and data of the additive manufacturing process;

[0024] Image processing and signal analysis module, used to process the data collected by the multimodal sensor module and identify defects;

[0025] Offline detection module, used to provide high-precision detection results required for calibration and model optimization;

[0026] Defect diagnosis and control decision module, including knowledge rule base, fuzzy controller or expert system, used to generate control instructions;

[0027] The real-time execution module is used to adjust parameters such as laser power, scanning speed, wire feed rate and trajectory path according to control instructions to achieve defect suppression and compensation.

[0028] The present invention also provides a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the closed-loop self-adjustment method for online defect identification and real-time feedback in a multi-beam high-power laser additive manufacturing process of titanium alloy components based on intelligent perception is executed.

[0029] The present invention also provides a computer device, which includes a memory and a processor, wherein the memory stores a computer program, and when the processor runs the computer program stored in the memory, the processor executes any one of the above-mentioned closed-loop self-adjustment methods for online identification and real-time feedback of defects in the multi-beam high-power laser additive manufacturing process of titanium alloy components based on intelligent perception.

[0030] The beneficial effects of the present invention are:

[0031] 1. The present invention proposes a closed-loop self-adjusting method for online defect identification and real-time feedback in the multi-beam high-power laser additive manufacturing process of titanium alloy components based on intelligent perception, aiming to solve key problems such as insufficient defect identification capabilities and lagging quality control in the current additive manufacturing process. The method is built on a multimodal sensing architecture, integrating an infrared thermal imager, a melt pool camera and a spectral sensor to achieve simultaneous monitoring of the melt pool thermal field, radiation characteristics and surface morphology of the deposited layer. Through traditional image processing and signal analysis methods, typical defects such as pores, cracks, and lack of fusion are extracted and identified; and offline high-precision detection methods such as industrial CT and metallographic analysis are combined for identification and calibration. The identification results are fed back to the central control platform in real time, and the cause is automatically diagnosed according to the defect type and distribution characteristics, the control strategy is matched, and the process parameters such as laser power and scanning path are dynamically adjusted to achieve defect suppression and online repair. The closed-loop mechanism supports system adaptive iterative optimization to improve forming quality and reliability.

[0032] Furthermore, the present invention achieves the following effects:

[0033] (1) Realize online accurate identification and highly reliable real-time feedback of typical defects in the laser additive manufacturing process;

[0034] (2) Maintaining the system’s interpretability and engineering portability through traditional image processing and signal analysis methods;

[0035] (3) Improve the reliability of the control strategy by combining closed-loop feedback with offline calibration;

[0036] (4) It can adapt to multi-beam laser high-power forming scenarios and meet the manufacturing requirements of high density, high strength and high reliability of titanium alloy components in the aerospace field.

[0037] The present invention can ultimately be widely used in the high-quality additive manufacturing process of large complex titanium alloy components for aerospace, providing intelligent monitoring and process assurance support for the stable performance of components. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following is a brief introduction to the drawings required for use in the specific embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0039] Figure 1 This is a flow chart of the closed-loop self-adjustment method for online defect identification and real-time feedback in the multi-beam high-power laser additive manufacturing process proposed by the present invention. DETAILED DESCRIPTION

[0040] In the following description, specific details such as specific system structures and technologies are provided for the purpose of illustration rather than limitation, so as to provide a thorough understanding of the embodiments of the present application. However, it should be clear to those skilled in the art that the present application can also be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to avoid unnecessary details that obscure the description of the present application.

[0041] The specific embodiments of the present invention are further described below in conjunction with the accompanying drawings. The following embodiments will help those skilled in the art further understand the present invention, but are not intended to limit the present invention in any form. It should be noted that those skilled in the art may make various changes and improvements without departing from the scope of the present invention, and these are all within the scope of protection of the present invention.

[0042] Implementation Method 1: Combination Figure 1 To describe this embodiment, the purpose of the present invention is to provide a closed-loop self-adjustment method for online identification and real-time feedback of defects in the multi-beam high-power laser additive manufacturing process of titanium alloy components based on intelligent perception, which aims to solve the problems of insufficient extraction of defect feature information and difficulty in online evaluation and feedback control of printing quality in the current laser additive manufacturing process, and to improve the forming quality and reliability of complex titanium alloy components (such as large suspensions, aerospace warhead casings, etc.).

[0043] like Figure 1 As shown, the method specifically includes the following steps:

[0044] Step S1: Online monitoring and data acquisition: A multi-source sensing system is used to synchronously collect the thermal field, light radiation, and morphological characteristics of the molten pool during the laser additive manufacturing process to obtain the original physical data of the deposited layer;

[0045] Step S2: Defect feature extraction and identification: Defect identification is performed on the original physical data of the sedimentary layer to extract the type, size, location and distribution information of the sedimentary layer defects;

[0046] Step S3: Offline detection data calibration: Perform offline detection on the sample to obtain the real metallurgical defect information inside the deposited layer of the sample, compare it with the identification result of step S2, and correct the defect information;

[0047] Step S4: Defect diagnosis and feedback control: The central computing and control platform diagnoses the cause based on the corrected defect information and matches the process control strategy;

[0048] Step S5: Real-time process parameter adjustment: Adjust process parameters based on the control strategy to complete online repair or suppression of typical defects;

[0049] Step S6: Closed-loop adjustment and continuous optimization: Compare the control results with the offline detection data of the final workpiece, and achieve adaptive closed-loop dynamic adjustment by optimizing the control strategy.

[0050] Implementation 2: This implementation is a specific description of the closed-loop self-adjustment method for online defect identification and real-time feedback in the multi-beam high-power laser additive manufacturing process of titanium alloy components based on intelligent perception proposed in the above implementation.

[0051] Step S1: using a multi-source sensing system to synchronously collect the thermal field, light radiation, and morphological characteristics of the molten pool during the laser additive manufacturing process to obtain the original physical data of the deposited layer;

[0052] Specifically:

[0053] The intelligent perception system proposed in this embodiment is built on a multimodal sensing architecture. By integrating multiple sensing methods, including an infrared thermal imager, a melt pool camera, and a spectral sensor, it simultaneously monitors the melt pool's thermal field, optical radiation, and the resulting surface topography during laser additive manufacturing. The raw data acquired by the system includes multi-dimensional information such as the temperature distribution, profile changes, and radiation energy of the deposited layer, providing a comprehensive data foundation for subsequent defect identification and diagnosis.

[0054] In practical application, the system first integrates multimodal sensing equipment on the additive manufacturing platform to achieve online perception of the entire manufacturing process. The sensor deployment includes infrared thermal imagers to monitor temperature distribution changes in the molten pool and its surrounding areas, industrial cameras to capture images of the deposited layer surface, and spectral sensors to detect the optical radiation characteristics generated by the interaction between the laser and the material. Data from these sensors is collected synchronously to build a multi-source fusion raw perception data platform.

[0055] Step S2: Defect feature extraction and identification: Defect identification is performed on the original physical data of the sedimentary layer to extract the type, size, location and distribution information of the sedimentary layer defects;

[0056] Specifically:

[0057] During the defect feature extraction and identification phase, this implementation utilizes traditional image processing and signal analysis methods, including edge detection, gray-level co-occurrence matrix, Fourier transform, and time-frequency analysis, to extract the type, size, location, and distribution characteristics of potential defects in the deposited layer. This method can identify typical defect types, including pores, cracks, lack of fusion holes, inclusions, edge collapse, and layer thickness fluctuations.

[0058] Step S3: Offline detection data calibration: Perform offline detection on the sample to obtain the real metallurgical defect information inside the deposited layer of the sample, compare it with the identification result of step S2, and correct the defect information;

[0059] Specifically:

[0060] To ensure online recognition accuracy, this implementation also incorporates an offline calibration mechanism. This involves analyzing representative samples using high-precision methods such as industrial CT scanning, metallographic analysis, fracture observation, and scanning electron microscopy to determine the distribution of metallurgical defects and their true structural state. Online recognition results are then compared with this high-precision data to refine threshold parameters, recognition logic, and control strategies, ensuring the accuracy of the online system.

[0061] Step S4: Defect diagnosis and feedback control: The central computing and control platform diagnoses the cause based on the corrected defect information and matches the process control strategy;

[0062] Specifically:

[0063] Identified defect information is transmitted in real time to a central computing and control platform, which includes a defect-process causal rule library, a process history data storage module, a statistical analysis unit, and an instruction generation module. Based on the identified defect types and distribution characteristics, combined with historical data and control rules, the platform can diagnose the cause and match appropriate control strategies.

[0064] Step S5: Real-time process parameter adjustment: Adjust process parameters based on the control strategy to complete online repair or suppression of typical defects;

[0065] Specifically:

[0066] The feedback control section proactively intervenes in defects and enables in-process repair by adjusting laser power, scanning speed, wire feed rate, and trajectory in real time. This allows for process operations such as melt pool shaping, localized remelting, and path redrawing. Specifically for typical metallurgical defects such as edge collapse or partial lack of fusion, the system can trigger selective rescanning or energy boost strategies to promptly suppress defect development during the manufacturing process.

[0067] Step S6: Closed-loop adjustment and continuous optimization: Compare the control results with the offline detection data of the final workpiece, and achieve adaptive closed-loop dynamic adjustment by optimizing the control strategy.

[0068] Specifically:

[0069] The closed-loop adjustment mechanism continuously compares the online control results with the offline inspection results of the final workpiece to extract the quality of the control effect, and optimizes the process parameters and identification strategy of the next cycle based on rule iteration and parameter correction methods, thereby realizing dynamic adaptive optimization of the system.

[0070] In summary, the closed-loop self-adjusting method for online defect identification and real-time feedback in the multi-beam high-power laser additive manufacturing process of titanium alloy components proposed in this embodiment is based on intelligent perception, aiming to solve key problems such as insufficient defect identification capabilities and lagging quality control in the current additive manufacturing process. This method is built on a multimodal sensing architecture, integrating an infrared thermal imager, a melt pool camera and a spectral sensor to achieve simultaneous monitoring of the melt pool thermal field, radiation characteristics and surface morphology of the deposited layer. Through traditional image processing and signal analysis methods, typical defects such as pores, cracks, and lack of fusion are extracted and identified; and combined with offline high-precision detection methods such as industrial CT and metallographic analysis for identification and calibration. The identification results are fed back to the central control platform in real time, and the cause is automatically diagnosed according to the defect type and distribution characteristics, the control strategy is matched, and the process parameters such as laser power and scanning path are dynamically adjusted to achieve defect suppression and online repair. The closed-loop mechanism supports system adaptive iterative optimization to improve forming quality and reliability.

[0071] Implementation method 3: This implementation method is to provide a practical application of the closed-loop self-adjustment method for online defect identification and real-time feedback in the multi-beam high-power laser additive manufacturing process of titanium alloy components based on intelligent perception proposed in the above implementation method;

[0072] In practical application, the system first integrates multimodal sensing equipment on the additive manufacturing platform to achieve online perception of the entire manufacturing process. The sensor deployment includes infrared thermal imagers to monitor temperature distribution changes in the molten pool and its surrounding areas, industrial cameras to capture images of the deposited layer surface, and spectral sensors to detect the optical radiation characteristics generated by the interaction between the laser and the material. Data from these sensors is collected synchronously to build a multi-source fusion raw perception data platform.

[0073] Based on sensory data acquisition, this implementation utilizes a variety of traditional image processing and signal analysis methods to rapidly analyze and identify defects. For example, irregular temperature anomalies in thermal images can be extracted through edge enhancement and binary segmentation to reveal potential holes or signs of unfused deposits. Deposition surface images can be analyzed using gray-level co-occurrence matrices and frequency domain analysis to identify uneven thickness or surface irregularities in the formed layer. Spectral line drift and intensity fluctuations in spectral data can reveal localized strong material evaporation and compositional instability. These processing methods are based on interpretable physical models, avoiding the use of deep learning black-box mechanisms and ensuring the traceability and stability of identification results.

[0074] To further improve recognition accuracy, the system also incorporates offline testing as a calibration tool. Metallographic analysis, X-ray scanning, and industrial CT imaging are performed on representative manufactured samples to obtain true metallurgical defect information within the deposited layer. This high-precision inspection data is then compared with online recognition results to optimize the recognition model's parameter settings, such as adjusting the discrimination thresholds for different defect types and correcting redundant or omitted feature extraction rules.

[0075] After a defect is accurately identified, the results are transmitted in real time to the central control platform. This platform, embedded with a process defect causal rule library and historical data analysis module, enables rapid diagnosis of the root cause of the defect and matches the corresponding control strategy based on statistical laws, process knowledge, and logical judgment. For example, if an abnormal increase in pore density and drastic temperature fluctuations are detected, the system can determine that the laser power or wire feed rate is unstable, triggering the relevant control response.

[0076] The control response primarily involves adjusting process parameters. The system can automatically adjust the laser power curve, change the scanning speed, optimize the path trajectory, or perform remelting and rescanning operations in localized areas to ensure that identified defects are effectively suppressed or repaired. This control process is not performed in isolation but forms a closed-loop feedback mechanism with subsequent sensor detection. The new state of the deposited layer is again monitored, identified, and diagnosed, and the control strategy is iteratively updated based on the feedback, thereby achieving adaptive system operation.

[0077] Ultimately, this invention achieves a closed-loop operating model of "perception-recognition-feedback-adjustment," with continuous self-optimization capabilities. As manufacturing tasks accumulate, the system automatically records the results of each recognition and adjustment, incorporating them into the iterative control rule process. This allows the entire platform to continuously improve stability and accuracy over the long term, adapting to the requirements of diverse component types and complex manufacturing paths.

[0078] Embodiment 4: This embodiment further provides an intelligent perception system implemented by a closed-loop self-adjustment method for online defect identification and real-time feedback in a multi-beam high-power laser additive manufacturing process of a titanium alloy component based on intelligent perception as described in the above embodiment. The system includes:

[0079] A multimodal sensor module for collecting infrared thermal images, visible light images, and contour data of the additive manufacturing process;

[0080] Image processing and signal analysis module, used to process the data collected by the multimodal sensor module and identify defects;

[0081] Offline detection module, used to provide high-precision detection results required for calibration and model optimization;

[0082] Defect diagnosis and control decision module, including knowledge rule base, fuzzy controller or expert system, used to generate control instructions;

[0083] The real-time execution module is used to adjust parameters such as laser power, scanning speed, wire feed rate and trajectory path according to control instructions to achieve defect suppression and compensation.

[0084] Implementation method five. This implementation method provides a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, a closed-loop self-adjustment method for online identification and real-time feedback of defects in a multi-beam high-power laser additive manufacturing process of titanium alloy components based on intelligent perception is executed.

[0085] Implementation method six. This implementation method provides a computer device, which includes a memory and a processor, wherein a computer program is stored in the memory. When the processor runs the computer program stored in the memory, the processor executes a closed-loop self-adjustment method for online identification and real-time feedback of defects in a multi-beam high-power laser additive manufacturing process of titanium alloy components based on intelligent perception as described in any one of the above implementation methods.

[0086] A computer device is provided in this embodiment. The hardware device of this part is a general model and is not shown in the form of a diagram. The system includes a processor and a memory, wherein the processor and the memory can be connected via a bus or other means. The memory is a non-transient computer-readable storage medium that can be used to store non-transient software programs, non-transient computer executable programs and modules, and corresponding program instructions / modules. The processor executes various functional applications and data processing of the processor by running the non-transient software programs, instructions and modules stored in the memory, so as to realize the closed-loop self-adjustment method and steps for online identification and real-time feedback of defects in the multi-beam high-power laser additive manufacturing process of titanium alloy components based on intelligent perception in the above method embodiment.

[0087] Through the above-mentioned implementation methods, the present invention not only effectively improves the forming quality of titanium alloy additive components, but also greatly reduces the rework rate and inspection costs during the manufacturing process, showing outstanding application prospects in the manufacturing of aerospace components with large size, high complexity and high reliability requirements.

[0088] The foregoing description is merely an embodiment of the present invention and is not intended to limit the present invention. Those skilled in the art will readily appreciate that the present invention is susceptible to various modifications and variations. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present invention shall be included within the scope of the claims.

Claims

1. A closed-loop self-adjustment method for online defect identification and real-time feedback in the additive manufacturing process of titanium alloy components based on intelligent perception, characterized in that: The method is: S1: A multi-source sensing system is used to synchronously collect the thermal field, optical radiation, and morphological characteristics of the molten pool during laser additive manufacturing to obtain the original physical data of the deposited layer; S2: Defect identification is performed on the original physical data of the sedimentary layer to extract the type, size, location and distribution information of the sedimentary layer defects; S3: Perform offline testing on the sample to obtain the true metallurgical defect information inside the deposited layer of the sample, compare it with the identification result of step S2, and correct the defect information; S4: The central computing and control platform diagnoses the cause based on the corrected defect information and matches the process control strategy; S5: Adjust process parameters based on control strategies to complete online repair or suppression of typical defects; S6: Compare the control results with the offline detection data of the final workpiece, and achieve adaptive closed-loop dynamic adjustment by optimizing the control strategy.

2. The closed-loop self-adjustment method for online defect identification and real-time feedback in the additive manufacturing process of titanium alloy components based on intelligent perception according to claim 1 is characterized in that: The multi-source perception system includes infrared thermal imagers, industrial cameras, and spectral sensors; Infrared thermal imagers are used to monitor temperature distribution changes in the molten pool and its surrounding areas; Industrial cameras are used to obtain images of the sediment surface; Spectral sensors are used to sense the optical radiation characteristics generated when lasers interact with materials.

3. The closed-loop self-adjustment method for online defect identification and real-time feedback in the additive manufacturing process of titanium alloy components based on intelligent perception according to claim 1 is characterized in that: Defect identification uses grayscale threshold method, edge detection method, image contour analysis or spectrum energy analysis.

4. The closed-loop self-adjustment method for online defect identification and real-time feedback in the additive manufacturing process of titanium alloy components based on intelligent perception according to claim 1 is characterized in that: Offline testing includes industrial CT scanning, metallographic analysis, metal fracture observation or scanning electron microscopy imaging.

5. The closed-loop self-adjustment method for online defect identification and real-time feedback in the additive manufacturing process of titanium alloy components based on intelligent perception according to claim 1 is characterized in that: Defect types include cracks, pores, metallurgical inclusions, edge collapse, uneven thickness of the formed layer or abnormal surface undulations.

6. The closed-loop self-adjustment method for online defect identification and real-time feedback in the additive manufacturing process of titanium alloy components based on intelligent perception according to claim 1 is characterized in that: The central computing and control platform is embedded with a defect type and process causal relationship rule library, a process parameter historical data storage module, a defect statistical analysis module, and a control instruction generation module.

7. The closed-loop self-adjustment method for online defect identification and real-time feedback in the additive manufacturing process of titanium alloy components based on intelligent perception according to claim 1 is characterized in that: The process control strategies include but are not limited to one or more of the following: dynamic adjustment of laser power, local switching of wire feeding modules, path redivision, and selective remelting.

8. An intelligent perception system implemented by the method according to any one of claims 1 to 7, characterized in that: include: A multimodal sensor module for collecting infrared thermal images, visible light images, and contour data of the additive manufacturing process; Image processing and signal analysis module, used to process the data collected by the multimodal sensor module and identify defects; Offline detection module, used to provide high-precision detection results required for calibration and model optimization; Defect diagnosis and control decision module, including knowledge rule base, fuzzy controller or expert system, used to generate control instructions; The real-time execution module is used to adjust parameters such as laser power, scanning speed, wire feed rate and trajectory path according to control instructions to achieve defect suppression and compensation.

9. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a computer program, which, when executed by a processor, executes the closed-loop self-adjustment method for online defect identification and real-time feedback in the additive manufacturing process of titanium alloy components based on intelligent perception as described in any one of claims 1 to 7.

10. A computer device, characterized in that: The device includes a memory and a processor, wherein a computer program is stored in the memory. When the processor runs the computer program stored in the memory, the processor executes the closed-loop self-adjustment method for online identification and real-time feedback of defects in the additive manufacturing process of titanium alloy components based on intelligent perception as described in any one of claims 1 to 7.

Images