A method for monitoring and adaptive control of powder densification state in hot-press sintering equipment

Abstract

The application discloses a kind of hot-pressing sintering equipment powder densification state monitoring and self-adaptive control method, first in the powder forming chamber periphery of continuous hot-pressing sintering equipment, construct the layered monitoring system consisting of direct perception layer, indirect compensation layer and global environment layer by multiple types of sensors;On the basis of redundant measuring point, measuring point selection strategy is established, and measuring point is dynamically activated according to different stages of sintering process;Time synchronization and preprocessing are carried out on the collected multi-source monitoring data, and the oscillation loading main frequency, oscillation amplitude and oscillation energy characteristics of structural vibration response signal are extracted;Densification state estimation model is constructed, and the densification state inside powder compact that cannot be directly measured is reconstructed;The operating parameters of inductive heating system and servo hydraulic oscillation loading system are adjusted.The method can realize the densification state perception of compact in powder sintering process, and provide basis and method for sintering process parameter regulation.

Description

Technical Field

[0001] This invention relates to the field of monitoring and control technology for powder metallurgy forming equipment, specifically to a method for monitoring and adaptive control of powder densification state in continuous hot pressing sintering equipment. Background Technology

[0002] Continuous hot pressing sintering is one of the important forming processes in the field of powder metallurgy for preparing high-density materials, and it is widely used in the manufacturing of cemented carbide tool blanks and high-performance structural materials. Compared with conventional sintering processes, vacuum continuous hot pressing sintering equipment can promote the plastic deformation and diffusion bonding of powder particles under the combined action of high temperature and pressure, enabling the powder blank to complete the densification process in a shorter time, thereby obtaining sintered products with high density and good mechanical properties. With the continuous improvement of the performance requirements of powder metallurgy materials, continuous hot pressing sintering equipment is gradually developing towards high temperature, high vacuum, and multi-process continuous operation, and is playing an increasingly important role in the preparation of high-performance materials such as tool blanks.

[0003] In the continuous hot pressing sintering process of vacuum powder, the densification behavior of the powder blank is typically influenced by the coupling effects of multiple physical fields, including temperature, loading path, and vacuum environment. As the sintering process progresses, powder particles gradually come into contact, rearrange, and undergo plastic deformation. The diffusion bonding between particles gradually strengthens, and the internal pore structure of the blank continuously evolves and gradually closes, ultimately forming a sintered product with a certain density and mechanical properties. During this process, the closure process and spatial distribution of the pores within the blank have a significant impact on the final density, microstructure, and mechanical properties of the sintered product. Therefore, effective monitoring and control of the densification behavior during sintering is of great importance for improving product quality.

[0004] In some continuous hot pressing sintering equipment, oscillatory loading is introduced to further improve the rearrangement and densification process of powder particles. By superimposing oscillatory stresses of a certain frequency and amplitude during loading, the powder particles undergo more thorough rearrangement and contact under periodic stress, thereby promoting the closure of the pore structure and improving the densification efficiency of the material. However, under oscillatory loading conditions, the stress state and densification evolution process inside the green body during sintering become more complex. How to effectively monitor the sintering state under oscillatory loading and achieve reasonable adjustment of process parameters has become an important issue in equipment operation and process optimization.

[0005] Existing vacuum powder hot pressing sintering equipment implements process control based on pre-set heating curves and loading paths, using directly measurable process parameters such as temperature, loading pressure, and mold displacement as the main monitoring basis to maintain stable operation of the sintering process. However, during the actual sintering process, the evolution of the pore structure inside the green body is under high temperature and high vacuum conditions, making it difficult to directly deploy conventional sensors inside the green body and to measure the pore closing process in real time. Therefore, existing equipment struggles to obtain crucial internal state information that directly reflects the degree of powder particle rearrangement and the pore closing state during operation.

[0006] In the absence of real-time feedback on the internal densification state, the sintering process typically relies on experience or preset process parameters for open-loop control. When heating and loading conditions do not match the densification process within the green body, premature densification of the surface area can occur while the internal pores are not fully closed. As loading stress continues to be applied, the unclosed pores may gradually evolve into closed residual pore structures, inhibiting subsequent diffusion densification and thus affecting the uniformity of the material's internal structure. Furthermore, under continuous production conditions, due to fluctuations in raw material powder characteristics and equipment operating conditions, the lack of effective monitoring and feedback adjustment of the internal densification state of the green body can exacerbate the problem of insufficient local density in sintered products, further leading to a decrease in material performance consistency.

[0007] Therefore, without compromising the sealing of the sintering environment and the stability of continuous equipment operation, how to obtain multi-source state information during the sintering process through reasonable multi-sensor monitoring methods, and how to effectively reconstruct the state of the pore closure process inside the powder blank that cannot be directly measured based on the monitoring data, while adaptively adjusting the equipment heating power output and loading path according to the obtained densification state, thereby optimizing the densification evolution process in the powder sintering process, reducing the formation of closed residual pores, and improving the consistency of sintered product quality, is a technical problem that urgently needs to be solved by those skilled in the art. Summary of the Invention

[0008] This invention addresses the problem of insufficient local density and excessively high closed porosity caused by the difficulty in directly monitoring the densification state inside the green body during oscillatory loading hot pressing sintering. It proposes a method for sensing and adaptively controlling the densification state of powder sintering under oscillatory loading conditions. By constructing a multi-layer monitoring system and extracting oscillatory loading response characteristics, the densification state inside the green body is reconstructed. Based on the deviation in the densification state, the heating system and the oscillatory loading system are adjusted, thereby improving the sintering densification process and enhancing product quality consistency.

[0009] To achieve the above objectives, the present invention employs the following technical solution:

[0010] Step 1) Multiple types of sensors are deployed around the powder forming chamber of the continuous hot pressing sintering equipment to construct a layered monitoring system consisting of a direct sensing layer, an indirect compensation layer, and a global environment layer. The direct sensing layer is used to collect directly obtainable process parameter information such as temperature, displacement, loading, and vacuum. The indirect compensation layer is used to obtain structural response information related to the internal state of the billet. The global environment layer is used to obtain equipment operating environment parameters. Multiple monitoring points are deployed at key monitoring locations to form monitoring redundancy and obtain multi-source monitoring data characterizing the state of the sintering process.

[0011] Step 2) Based on the redundantly deployed measurement points, a measurement point selection strategy is constructed, different combinations of measurement points are analyzed, the Pareto multi-objective monkey swarm algorithm is used to solve the combination of monitoring measurement points, and the measurement points are dynamically activated according to different stages of the sintering process.

[0012] Step 3) Collect multi-source monitoring data according to the measurement point activation strategy, and perform time synchronization processing and preprocessing on the collected data.

[0013] Step 4) Extract features from the collected structural vibration response signals to obtain the oscillation loading main frequency, oscillation amplitude and oscillation energy features, and fuse the oscillation features with temperature, mold displacement and loading force data.

[0014] Step 5) Construct a densification state estimation model based on the fused multi-source monitoring data, reconstruct the densification state inside the powder blank that cannot be directly measured, and obtain state variables characterizing the degree of densification of the blank.

[0015] Step 6) Generate process control instructions based on the deviation between the densification state variables and the target densification trajectory, and adjust the operating parameters of the induction heating system and the servo hydraulic oscillation loading system.

[0016] Furthermore, in step 1, the direct sensing layer collects directly obtainable process parameter information such as mold temperature, mold temperature gradient, mold displacement, and oscillation loading force during the sintering process by deploying temperature sensors, displacement sensors, and load sensors in the mold system and oscillation loading system.

[0017] Furthermore, in step 1, the indirect compensation layer obtains the structural vibration response signal generated by oscillating loading by arranging an acceleration sensor outside the mold base or loading structure, and obtains the acoustic emission signal related to the particle rearrangement and pore evolution inside the blank by arranging an acoustic emission sensor in the mold system.

[0018] Furthermore, in step 1, the global environment layer obtains chamber vacuum level and inlet / outlet pressure information by deploying vacuum sensors and pressure sensors in the multi-chamber vacuum system, and obtains hydraulic oil pressure and heating power data by deploying pressure sensors and power meters in the oscillation loading system and induction heating system, respectively.

[0019] Furthermore, the measurement point selection strategy in step 2 is to optimize the combination of monitoring measurement points using the Pareto multi-objective monkey swarm algorithm. The specific method is as follows:

[0020] First, a set of monitoring points is constructed based on the multiple monitoring points deployed in step 1, and an evaluation index system for the monitoring point combination is established. The evaluation index includes monitoring coverage, monitoring information gain, monitoring reliability, spatial environmental constraints, and signal independence.

[0021] Secondly, a multi-objective optimization model is constructed with the evaluation index as the optimization objective, and the Pareto multi-objective monkey swarm algorithm is used to search and optimize the combination of monitoring points to obtain a set of monitoring point combinations that satisfy different optimization objectives.

[0022] Finally, based on the process characteristics of different stages of the sintering process, the corresponding monitoring point combination is selected from the set of monitoring point combinations and dynamically activated to achieve adaptive acquisition of sintering process monitoring data.

[0023] Furthermore, the method for collecting and processing multi-source monitoring data in step 3 is as follows:

[0024] First, based on the monitoring point activation strategy obtained in step 2, data is collected from the activated monitoring points to obtain multi-source monitoring data such as temperature, mold displacement, loading force, vacuum degree, structural vibration response, and acoustic emission.

[0025] Secondly, the collected multi-source monitoring data is processed for time synchronization, so that the data collected by different types of sensors are aligned under a unified time reference.

[0026] Finally, the synchronized multi-source monitoring data is preprocessed, including signal filtering, outlier removal, and data normalization, to obtain stable and reliable monitoring data.

[0027] Furthermore, the feature extraction method for the structural vibration response signal in step 4 is as follows:

[0028] First, frequency domain analysis is performed on the obtained structural vibration response signal. The dominant frequency component during the oscillating loading process is obtained by fast Fourier transform, thus obtaining the dominant frequency characteristics of the oscillating loading.

[0029] Secondly, time-domain analysis is performed on the structural vibration response signal. By statistically analyzing the changes in the vibration signal amplitude, the oscillation amplitude characteristics are obtained, which characterize the oscillation loading intensity.

[0030] Finally, the energy characteristics of the oscillation signal are calculated based on the time and frequency domain information of the structural vibration response signal. The oscillation main frequency characteristics, oscillation amplitude characteristics, and oscillation energy characteristics are then fused with temperature, mold displacement, and loading force data to form a feature dataset for densification state reconstruction.

[0031] Furthermore, the method for establishing the densification state estimation model in step 5 is as follows:

[0032] First, a set of state variables for the sintering process is constructed based on the feature dataset. The state variables include mold temperature, loading force, mold displacement, oscillation characteristics, and acoustic emission characteristics.

[0033] Secondly, a mapping model between the densification state during the sintering process and multi-source monitoring characteristics is established to estimate the densification state inside the powder compact that cannot be directly measured. Its expression is as follows:

[0034]

[0035] In the formula: Indicates the time of the billet Compacted state variables, Indicates the mold temperature. Indicates the applied force. Indicates mold displacement. Indicates oscillatory loading characteristics, This indicates the characteristics of acoustic emission.

[0036] Finally, the densification state estimation model is trained and updated online based on the multi-source monitoring feature data, thereby realizing the real-time reconstruction of the densification state of the green body during sintering. The densification state variables are used to characterize the changes in the relative density or porosity of the green body.

[0037] Furthermore, the process control method in step 6 is as follows:

[0038] First, the deviation between the current densification state and the target state is calculated by comparing the densification state variables of the billet with the preset target densification trajectory.

[0039] Secondly, process control instructions are generated based on the densification state deviation, and the process parameters that need to be adjusted are determined, including heating temperature, loading force, and oscillation loading frequency or amplitude.

[0040] Finally, the process control command is sent to the induction heating system and the servo hydraulic oscillation loading system. By adjusting the heating power, loading pressure and oscillation loading parameters, the adaptive control of the sintering process parameters is achieved.

[0041] Compared with the prior art, the present invention has the following beneficial effects:

[0042] (1) This invention addresses the problem of insufficient monitoring methods in vacuum powder continuous hot pressing sintering equipment under high temperature, high vacuum and space-constrained conditions. It constructs a layered monitoring system consisting of a direct sensing layer, an indirect compensation layer and a global environment layer, and deploys multiple types of sensors at key monitoring locations to acquire multi-source monitoring data such as temperature, displacement, loading, vacuum and structural vibration. At the same time, based on the multi-point deployment, the Pareto multi-objective monkey swarm algorithm is introduced to optimize the combination of monitoring points, and the dynamic activation of monitoring points is carried out according to different stages of the sintering process. This enables the monitoring system to reduce the redundancy of monitoring points while ensuring coverage of key parameters, and improves the stability and effectiveness of monitoring data acquisition under complex working conditions.

[0043] (2) This invention introduces the vibration response signal of the mold base structure, extracts the main frequency, amplitude and energy features during the oscillation loading process, and integrates them with process parameters such as temperature, mold displacement and loading force to establish a densification state estimation model. This enables online reconstruction of the densification state inside the powder blank, which cannot be directly measured, so that the sintering process under oscillation loading conditions can be transformed from monitoring a single process parameter to process state perception based on multi-source feature fusion.

[0044] (3) Based on the densification state of the billet, the present invention generates process control instructions according to the deviation between the densification state and the target densification trajectory, and coordinates the induction heating system and the servo hydraulic oscillation loading system. Through adaptive control of heating power, loading pressure and oscillation loading parameters, the sintering process can be dynamically adjusted according to the densification evolution state of the billet, thereby improving the stability and consistency of the sintering densification process of the tool billet and reducing the risk of insufficient local density and the formation of closed pores. Attached Figure Description

[0045] Figure 1 This is a schematic diagram of the vacuum powder continuous hot pressing sintering equipment in this invention;

[0046] Figure 2 This is a schematic diagram of the powder forming chamber and sensor layout in this invention;

[0047] Figure 3 This is the hierarchical monitoring architecture of the present invention;

[0048] Figure 4 This is a flowchart of the measurement point optimization and dynamic activation method of the present invention;

[0049] Figure 5 This is a flowchart of the oscillation loading feature extraction process of the present invention;

[0050] Figure 6This is a schematic diagram of the densification state reconstruction model during the sintering process of the present invention;

[0051] Figure 7 This is a schematic diagram of the closed-loop control structure for the sintering process of the present invention. Detailed Implementation

[0052] The present invention will now be described in further detail with reference to the accompanying drawings:

[0053] Example 1

[0054] like Figure 1 As shown, this embodiment uses a vacuum powder continuous hot pressing sintering equipment as the application object. The equipment includes a charging chamber 1, a central chamber 2, a powder forming chamber 3, a heat treatment chamber 4, and a material inlet / outlet chamber 5. The chambers are connected through the central chamber 2, and the material is automatically transferred between the chambers by a transfer robot 7. The equipment also includes a vacuum system 6 for maintaining a high vacuum environment inside the equipment.

[0055] In the powder forming process, the powder material first enters the equipment through the loading chamber 1 and is then transported to the powder forming chamber 3 by the transfer robot 7. Inside the powder forming chamber 3, the powder blank is heated and loaded by the oscillating loading system 8 and the induction heating system 9, thereby achieving the densification of the powder material. The mold system 10 is located inside the powder forming chamber 3 to contain the powder material and withstand the loading pressure.

[0056] After the powder forming is completed, the blank is transported to the heat treatment chamber 4 by the transfer robot 7 for subsequent heat treatment, and finally the material is loaded and unloaded through the loading and unloading chamber 5.

[0057] In this continuous hot pressing sintering process, the densification state of the powder blank is difficult to obtain directly. Therefore, it is necessary to use multi-source sensing information to monitor and reconstruct the state of the sintering process in real time in order to achieve effective control of the sintering process.

[0058] Step 1: Deployment of multi-source sensors and construction of a hierarchical monitoring system.

[0059] In this embodiment, multiple types of sensors are first deployed around the powder forming chamber 3 of the continuous hot pressing sintering equipment to obtain key process parameters and structural response information during the sintering process.

[0060] like Figure 2A displacement sensor 11 is installed at the oscillating loading system 8 to collect the displacement changes of the mold during the loading process; a load sensor 12 is installed at the oscillating loading system 8 to obtain the loading force applied to the mold system 10 during the sintering process; a vacuum sensor 13 is installed on the cavity wall of the powder forming chamber 3 to monitor the vacuum environment inside the equipment; a temperature sensor 14 is installed on the mold system 10 to collect the mold temperature and its changes; an acceleration sensor 15 is installed at the mold base of the mold system 10 to obtain the structural vibration response signal generated during the oscillating loading process; and an acoustic emission sensor 16 is installed on the mold system 10 to collect the acoustic emission signal generated by particle rearrangement and pore evolution during the sintering process.

[0061] like Figure 3 Based on the sensor deployment, a layered monitoring architecture is constructed, consisting of a direct sensing layer, an indirect compensation layer, and a global environment layer. The direct sensing layer collects readily available key process parameters, including mold temperature, mold displacement, loading force, and mold temperature gradient. The indirect compensation layer acquires structural response information related to the densification state of the billet, including vibration response signals and acoustic emission signals generated by oscillatory loading. The global environment layer acquires equipment operating environment information, including chamber vacuum parameters, heating power parameters, and hydraulic pressure signals. Through this layered monitoring architecture, multi-dimensional process information of the sintering process is collected, forming multi-source monitoring data for state reconstruction.

[0062] Step 2: Optimization and dynamic activation of monitoring points.

[0063] To improve the efficiency of monitoring information acquisition and reduce redundant data collection, the combination of monitoring points was optimized based on the redundant deployment of monitoring points, and the monitoring points were dynamically activated according to different stages of the sintering process.

[0064] like Figure 4 First, a candidate monitoring point set is constructed. The candidate monitoring point set consists of various sensor points deployed in step 1, including temperature sensor 14, displacement sensor 11, load sensor 12, vacuum sensor 13, acceleration sensor 15, and acoustic emission sensor 16.

[0065] Secondly, an evaluation index system for monitoring points is established to assess the impact of different combinations of monitoring points on the ability to characterize the sintering process state. The evaluation indexes include monitoring coverage, information gain, reliability, spatial constraints, and signal independence. Monitoring coverage evaluates the ability of the monitoring point combination to cover key process parameters; information gain measures the improvement in the ability to acquire state information by adding new monitoring points; reliability assesses the stability of the monitoring system under sensor failure or signal drift conditions; spatial constraints reflect the limitations imposed by equipment structure space and installation conditions on the layout of monitoring points; and signal independence describes the degree of signal correlation between different monitoring points.

[0066] Based on this, a multi-objective measurement point optimization model is constructed, and the above evaluation indicators are used as optimization objectives. The Pareto multi-objective monkey swarm algorithm is used to optimize the combination of monitoring measurement points, thereby obtaining the optimal combination of monitoring measurement points that satisfies the multi-objective constraints.

[0067] After obtaining the optimal combination of monitoring points, the monitoring points are dynamically activated according to the process stages of sintering. Specifically, based on the heating stage, holding stage, and oscillation loading stage in the sintering process, corresponding key monitoring points are selected for activation, so that the monitoring system focuses on acquiring relevant key status information at different stages.

[0068] Step 3: Multi-source monitoring data acquisition and preprocessing.

[0069] After obtaining the optimal combination of monitoring points and completing the dynamic activation of the monitoring points, the multi-source monitoring data collected from each monitoring point are uniformly collected and preprocessed to obtain the monitoring data sequence for subsequent feature extraction.

[0070] Specifically, the data acquisition module collects data from the various sensors activated in step 2, including mold temperature data collected by temperature sensor 14, mold displacement data collected by displacement sensor 11, loading force data collected by load sensor 12, chamber vacuum data collected by vacuum sensor 13, structural vibration signals collected by acceleration sensor 15, and acoustic emission signals collected by acoustic emission sensor 16.

[0071] Because the aforementioned multi-source monitoring data differ in sampling frequency, time scale, and signal format, the collected data were first time-synchronized to unify the time reference for various monitoring data. Subsequently, signal preprocessing was performed on the monitoring data, including filtering vibration and acoustic emission signals to reduce environmental noise interference, removing abnormal sampling points, and standardizing monitoring data with different dimensions.

[0072] Step 4: Oscillating loading feature extraction.

[0073] Feature extraction is performed on the structural response signals collected during the sintering process to obtain characteristic parameters that can characterize the oscillating loading state and the internal evolution process of the material.

[0074] like Figure 5 First, the oscillating loading response signal collected by the accelerometer 15 is acquired, and feature extraction is performed on this signal. Specifically, time-domain and frequency-domain analyses are performed on the vibration signal. A fast Fourier transform is used to perform spectral analysis on the vibration signal to extract the dominant frequency information during the oscillating loading process. Simultaneously, statistical analysis is performed on the amplitude changes of the vibration signal to obtain the amplitude characteristics of the oscillating loading. The oscillation energy characteristics are then calculated based on the vibration signal. Oscillation characteristic parameters used to characterize the oscillating loading state are obtained.

[0075] Simultaneously, feature extraction is performed on the acoustic emission signal collected by the acoustic emission sensor 16 to reflect the acoustic emission response generated by powder particle rearrangement and pore evolution during sintering. Specifically, the acoustic emission signal is filtered, and characteristic parameters of the acoustic emission signal are extracted to obtain acoustic emission characteristic data.

[0076] Through the above feature extraction process, oscillation feature parameters and acoustic emission feature parameters can be obtained, and together with process parameter data such as temperature, loading force and mold displacement, they constitute a feature dataset for reconstructing the densification state.

[0077] Step 5: Reconstruction of the densified state.

[0078] The densification state of the powder compact is reconstructed to characterize the internal state of the compact during sintering.

[0079] like Figure 6 The oscillation characteristic parameters and acoustic emission characteristic parameters obtained in step 4 are fused with the temperature data, loading force data and mold displacement data obtained in step 3 to construct a set of characteristic variables to describe the densification state of the green body during sintering.

[0080] Based on this, a densification state estimation model is established to describe the mapping relationship between multi-source characteristic parameters and the densification state of the billet. The densification state estimation model can be expressed as:

[0081]

[0082] In the formula: Indicates the time of the billet Compacted state variables, Indicates the mold temperature. Indicates the applied force. Indicates mold displacement. This represents the oscillation characteristic parameters extracted by the accelerometer 15. This represents the acoustic emission characteristic parameters extracted by the acoustic emission sensor 16. This represents the mapping relationship between multi-source feature parameters and compaction states.

[0083] In practical applications, the mapping relationship can be solved using a data-driven model, a mechanism model, or a mechanism-data fusion model based on multi-source monitoring data, thereby obtaining the estimation result of the compaction state of the billet.

[0084] Step six: Closed-loop control of the sintering process.

[0085] The sintering process parameters are adjusted according to the densification state to achieve closed-loop control of the sintering process.

[0086] like Figure 7 First, the compacted state of the green body obtained in step 5 is... Compacting trajectory with the preset target By comparing the current densification state with the target densification state, the deviation between the current densification state and the target densification state is calculated, thereby obtaining state deviation information.

[0087] Then, a process control command is generated based on the state deviation, and the control command is sent to the process execution system of the sintering equipment. The process execution system includes an induction heating system 9 and an oscillating loading system 8, wherein the induction heating system 9 is used to adjust the heating power during the sintering process, and the oscillating loading system 8 is used to adjust the loading pressure and the frequency or amplitude of the oscillation loading.

[0088] Under the action of the process execution system, the heating and loading conditions of the mold system 10 during the sintering process are adjusted, thereby changing the temperature field and stress state of the powder blank during the sintering process and causing the densification process of the blank to evolve towards the target densification trajectory.

[0089] During the above-mentioned control process, the state response generated during the sintering process is monitored in real time again by temperature sensor 14, displacement sensor 11, load sensor 12, acceleration sensor 15 and acoustic emission sensor 16, and then re-enters the data acquisition and processing process in step 3, thereby forming a closed-loop control mechanism for the sintering process based on multi-source monitoring information.

[0090] The above method enables real-time sensing of the densification state of the green body and adaptive control of process parameters during continuous hot pressing sintering, thereby improving the stability and consistency of the sintering process.

[0091] Example 2

[0092] This embodiment uses the continuous hot pressing and sintering process of WC-Co cemented carbide tool blanks as an application scenario to illustrate the multi-dimensional monitoring data acquisition and fusion method proposed in this invention.

[0093] In this embodiment, the following is adopted: Figure 1 The vacuum powder continuous hot pressing sintering equipment shown sintersects WC-Co powder into shape. The WC-Co powder first enters the equipment through the loading chamber 1, where it is transferred by the transfer robot 7 in the central chamber 2 and then sent to the powder forming chamber 3 for hot pressing sintering. After sintering, the blank is transported by the transfer robot 7 to the heat treatment chamber 4 for subsequent heat treatment, and the material is fed and discharged through the inlet / outlet chamber 5.

[0094] During the powder forming process, various types of sensors are deployed around the powder forming chamber 3 to obtain multi-source monitoring data of the sintering process. For example... Figure 2 A displacement sensor 11 is installed at the oscillation loading system 8 to collect mold displacement, a load sensor 12 is installed at the oscillation loading system 8 to collect loading force, a vacuum sensor 13 is installed at the cavity wall of the powder forming chamber 3 to collect the cavity vacuum degree, a temperature sensor 14 is installed on the mold system 10 to collect mold temperature, an acceleration sensor 15 is installed at the mold seat of the mold system 10 to collect oscillation loading response signal, and an acoustic emission sensor 16 is installed on the mold system 10 to collect acoustic emission signal generated during sintering.

[0095] Based on the sensor deployment, construct such as Figure 3 The layered monitoring architecture is shown below. The direct sensing layer acquires process parameters such as mold temperature, mold displacement, and loading force; the indirect compensation layer acquires structural response information related to the densification process within the preform, including vibration response signals generated by oscillating loading and acoustic emission signals; and the global environment layer acquires equipment operating environment information, including chamber vacuum, hydraulic oil pressure, and heating power.

[0096] Subsequently, according to Figure 4 The illustrated process optimizes the combination of monitoring points. First, a candidate set of monitoring points is constructed, and an evaluation index system for these points is established. This evaluation index includes monitoring coverage, information gain, reliability, spatial constraints, and signal independence. Based on these evaluation indexes, a multi-objective optimization model is constructed, and the Pareto multi-objective monkey swarm algorithm is used to optimize the combination of monitoring points, thereby obtaining the optimal combination for monitoring the sintering process. During the sintering process, the monitoring points are dynamically activated according to different process stages, enabling the monitoring system to focus on acquiring relevant key status information during the heating, holding, and oscillation loading stages.

[0097] After obtaining the optimal combination of measuring points, the multi-source data collected by each sensor is collected in a unified manner through the data acquisition module, and the collected data is preprocessed, including data time synchronization, signal filtering and data standardization, to obtain a multi-source monitoring data sequence with a unified time reference.

[0098] Furthermore, according to Figure 5 The process shown involves feature extraction. Frequency and time domain analyses are performed on the vibration signal acquired by the accelerometer 15. The dominant oscillation frequency information is extracted using a Fast Fourier Transform, and the oscillation amplitude characteristics are obtained through vibration signal amplitude statistics. Simultaneously, vibration energy characteristics are calculated to obtain oscillation characteristic parameters. The acoustic emission signal acquired by the acoustic emission sensor 16 is filtered, and acoustic emission characteristic parameters are extracted.

[0099] The above characteristics are fused with temperature, loading force, and mold displacement data, and then processed according to... Figure 6 The densification state estimation model shown reconstructs the densification state of the green body during sintering, and its expression is as follows:

[0100]

[0101] In the formula: Indicates the time of the billet Compacted state variables, Indicates the mold temperature. Indicates the applied force. Indicates mold displacement. Represents the oscillation characteristic parameters. This represents the characteristic parameters of acoustic emission.

[0102] After obtaining the densification state estimation results, according to Figure 7 The closed-loop control structure shown regulates the sintering process parameters. Specifically, it controls the current densification state. Compacted trajectory with preset target The two systems are compared, and process control instructions are generated based on the deviation between them. These instructions are then sent to the induction heating system 9 and the oscillation loading system 8 to adjust the heating power, loading pressure, and oscillation loading frequency or amplitude during the sintering process.

[0103] During the control process, the state response generated by the sintering process is monitored again by temperature sensor 14, displacement sensor 11, load sensor 12, acceleration sensor 15 and acoustic emission sensor 16, and re-enters the data acquisition and processing flow, thereby forming a closed-loop control of the sintering process based on multi-source monitoring data.

[0104] In summary, this invention constructs a layered monitoring architecture by deploying multiple types of sensors around the powder forming chamber of a continuous hot-pressing sintering equipment to collect temperature, loading, displacement, oscillation response, and acoustic emission signals during the sintering process. Based on redundant measuring points, a multi-objective optimization method is used to determine the optimal combination of monitoring points, and these points are dynamically activated according to the sintering process stages. Furthermore, feature extraction and fusion processing are performed on the multi-source monitoring data, and a densification state estimation model is established to reconstruct the densification state of the powder blank. Subsequently, closed-loop control of the sintering process parameters is implemented based on the densification state. Through this method, the densification state of the blank during continuous hot-pressing sintering can be characterized, providing a basis for process adjustment during the sintering process.

[0105] The above content is only for illustrating the technical concept of the present invention and should not be construed as limiting the scope of protection of the present invention. Any modifications made to the technical solution based on the technical concept proposed in this invention shall fall within the scope of the claims of this invention.

Claims

1. A method for monitoring and adaptively controlling the densification state of powder in a hot pressing sintering equipment, characterized in that, Includes the following steps: Step 1) Deploy multiple types of sensors around the powder forming chamber of the continuous hot pressing sintering equipment to construct a layered monitoring system consisting of a direct sensing layer, an indirect compensation layer, and a global environment layer, in order to obtain multi-source monitoring data of the sintering process; Step 2) Based on the redundantly deployed monitoring points, a monitoring point selection strategy is constructed. The Pareto multi-objective monkey swarm algorithm is used to optimize the combination of monitoring points, and the monitoring points are dynamically activated according to different stages of the sintering process. Step 3) Collect multi-source monitoring data according to the measurement point activation strategy, and perform time synchronization processing and preprocessing on the collected data; Step 4) Extract features from the collected structural vibration response signals to obtain the oscillation loading main frequency, oscillation amplitude, and oscillation energy features, and fuse the oscillation features with temperature, mold displacement, and loading force data; Step 5) Construct a densification state estimation model based on the fused multi-source monitoring data to reconstruct the densification state inside the powder blank that cannot be directly measured; Step 6) Generate process control instructions based on the deviation between the densification state and the target densification trajectory, and adjust the operating parameters of the induction heating system and the servo hydraulic oscillation loading system.

2. The method for monitoring and adaptive control of powder densification state in hot pressing sintering equipment according to claim 1, characterized in that: The hierarchical monitoring system in step 1 includes a direct sensing layer, an indirect compensation layer, and a global environment layer.

3. The method for monitoring and adaptive control of powder densification state in hot pressing sintering equipment according to claim 1, characterized in that: The direct sensing layer collects mold temperature, mold displacement, and oscillating loading force by deploying temperature sensors, displacement sensors, and load sensors in the mold system and oscillation loading system.

4. The method for monitoring and adaptive control of powder densification state in hot pressing sintering equipment according to claim 1, characterized in that: The indirect compensation layer obtains structural vibration response signals by deploying acceleration sensors outside the mold base or loading structure, and obtains acoustic emission signals by deploying acoustic emission sensors in the mold system.

5. The method for monitoring and adaptive control of powder densification state in hot pressing sintering equipment according to claim 1, characterized in that: The global environment layer obtains chamber vacuum and pressure information by deploying vacuum and pressure sensors in the vacuum system, and obtains hydraulic oil pressure and heating power by deploying pressure sensors and power meters in the oscillating loading system and induction heating system.

6. The method for monitoring and adaptive control of powder densification state in hot pressing sintering equipment according to claim 1, characterized in that: In step 2, the Pareto multi-objective monkey swarm algorithm is used to optimize the combination of monitoring points. The optimization objectives include monitoring coverage, monitoring information gain, monitoring reliability, spatial environmental constraints, and signal independence.

7. The method for monitoring and adaptive control of powder densification state in hot pressing sintering equipment according to claim 1, characterized in that: In step 4, the oscillation loading characteristics are extracted by performing frequency domain analysis and time domain analysis on the structural vibration response signal. Specifically, the oscillation loading main frequency characteristics are obtained by fast Fourier transform, the oscillation amplitude characteristics are obtained by vibration signal amplitude statistics, and the oscillation energy characteristics are calculated based on the vibration signal.

8. The method for monitoring and adaptive control of powder densification state in hot pressing sintering equipment according to claim 1, characterized in that: In step 5, a mapping model between multi-source monitoring characteristics and the densification state of the powder blank is established to estimate the densification state inside the powder blank that cannot be directly measured.

9. The method for monitoring and adaptive control of powder densification state in hot pressing sintering equipment according to claim 1, characterized in that: The densification state estimation model constructed in step 5 is used to characterize the change in relative density or porosity of the billet.

10. The method for monitoring and adaptive control of powder densification state in hot pressing sintering equipment according to claim 1, characterized in that: In step 6, process control instructions are generated based on the deviation between the densification state of the billet and the target densification trajectory, so as to adjust the heating power of the induction heating system and the loading pressure and oscillation loading frequency or amplitude of the servo hydraulic oscillation loading system.

Images