Friction solid phase additive manufacturing temperature field identification control method and system

Abstract

The invention discloses a friction solid phase additive manufacturing temperature field identification control method and system, and relates to the technical field of additive manufacturing, and the method comprises the steps: carrying out the spatial domain and time domain combined decoupling processing of real-time temperature field data through a model, and obtaining a spatial domain feature weight and a time domain feature weight; determining spatial difference data between the real-time temperature field and the reference temperature field; determining local abnormal disturbance candidate regions and disturbance characteristics corresponding to the regions according to the spatial difference data; determining a real-time disturbance abnormal source position and disturbance intensity according to the disturbance characteristics and the current process state parameters; determining an initial friction head state adjustment parameter according to the real-time disturbance abnormal source position and the disturbance intensity, and performing correction according to the head track of the friction head to obtain a corrected friction head state adjustment parameter; accurate recognition and dynamic control of the real-time temperature field in the friction solid-phase additive manufacturing process are achieved, and the process stability and the part forming quality are remarkably improved.

Description

Technical Field

[0001] This invention relates to the field of additive manufacturing technology, specifically to a method and system for temperature field identification and control in triboelectric solid-phase additive manufacturing. Background Technology

[0002] Friction solid-state additive manufacturing (FSM) is an emerging additive manufacturing process based on the principle of friction stir welding. It achieves layer-by-layer material deposition to ultimately form a solid structure through the rotation and feeding of a friction head. Compared with traditional additive manufacturing technologies such as fused deposition modeling (FDM) and selective laser melting (SLM), FSM offers advantages such as low heat input, high weld quality, and low manufacturing cost. In recent years, it has experienced rapid development and widespread application in fields such as aerospace, high-speed rail, and automobile manufacturing.

[0003] However, in actual production processes, the real-time temperature field of triboelectric solid-phase additive manufacturing often exhibits significant local anomalies. These anomalies lead to uneven temperature distribution in the processing area, and may even cause defects such as localized overheating or incomplete fusion of the material, directly affecting the mechanical properties and manufacturing precision of the final part. Currently, conventional temperature field monitoring and control methods mainly employ simple temperature sensors combined with fixed thresholds for anomaly monitoring. However, this method struggles to accurately identify the specific location and intensity of local anomalies, resulting in delayed or inaccurate adjustments to process parameters, making it difficult to meet the manufacturing requirements of high-performance parts.

[0004] In addition, existing temperature control schemes mostly rely on pre-set parameters and lack effective technical means to dynamically adjust the real-time status of the friction head. In practical applications, when complex disturbances occur, it can easily lead to instability in the processing or inconsistency in the performance of parts.

[0005] Therefore, in order to address the above problems, there is an urgent need for a temperature field identification and control method that can accurately identify local temperature anomalies during the triboelectric solid-phase additive manufacturing process in real time and can dynamically and accurately adjust the state parameters of the friction head, so as to improve the stability of the triboelectric solid-phase additive manufacturing process and product quality. Summary of the Invention

[0006] The purpose of this invention is to provide a method and system for identifying and controlling the temperature field in triboelectric solid-phase additive manufacturing, so as to solve the problems in the background art.

[0007] To achieve the above objectives, the present invention provides the following technical solution:

[0008] In a first aspect, the present invention provides a method for temperature field identification and control in triboelectric solid-state additive manufacturing, comprising:

[0009] S101: Obtain real-time temperature field data and current process state parameters during the triboelectric solid-phase additive manufacturing process; use a pre-built space-time temperature field coupling model to perform joint decoupling processing of the real-time temperature field data in the spatial and time domains to obtain spatial domain feature weights and time domain feature weights.

[0010] S102: Based on the real-time temperature field data and the spatial domain feature weights and time domain feature weights, determine the spatial difference data between the real-time temperature field and the reference temperature field; based on the spatial difference data, determine the candidate regions for local abnormal disturbances and the disturbance characteristics corresponding to each region;

[0011] S103: Determine the location and intensity of the real-time disturbance anomaly source based on the disturbance characteristics of the candidate region and the current process state parameters; determine the initial friction head state adjustment parameters based on the location and intensity of the real-time disturbance anomaly source, and correct them based on the friction head trajectory to obtain the corrected friction head state adjustment parameters;

[0012] S104: Adjust the feed trajectory of the friction head in real time according to the adjusted parameters of the friction head state; and dynamically update the space-time temperature field coupling model according to the updated real-time temperature field data for the next real-time temperature field data processing.

[0013] Furthermore, the joint spatial and temporal decoupling processing of the real-time temperature field data includes:

[0014] A spatial-temporal feature tensor is constructed based on historical temperature field data, and an improved three-dimensional convolutional attention autoencoder is used to determine the initial spatial feature weights of each spatial region and the initial temporal feature weights of each time step.

[0015] The temperature change rate of each spatial region is determined based on real-time temperature field data, and the initial spatial feature weights are adjusted using the temperature change rate to obtain the spatial domain feature weights.

[0016] The temperature change rate at each moment is determined based on real-time temperature field data, and the initial time feature weights are adjusted using the temperature change rate to obtain the time domain feature weights.

[0017] Furthermore, the adjustment of the initial spatial feature weights using the temperature change rate includes:

[0018] The temperature change sensitivity index for each spatial region is determined based on the rate of temperature change.

[0019] Calculate the probability of temperature anomalies in a spatial region based on the temperature change sensitivity index;

[0020] The initial spatial feature weights are adjusted by weighting the initial spatial feature weights based on the probability of temperature anomalies, and the spatial domain feature weights are determined.

[0021] Furthermore, adjusting the initial time feature weights using the rate of temperature change includes:

[0022] The temperature change sensitivity index is determined based on the rate of temperature change.

[0023] Determine the temperature anomaly trend coefficient based on the temperature change sensitivity index;

[0024] The initial time feature weights are adjusted by weighting the initial time feature weights based on the temperature anomaly trend coefficient, and the time domain feature weights are determined.

[0025] Furthermore, the process of determining candidate regions for local anomalous disturbances and the corresponding disturbance characteristics of each region includes:

[0026] Calculate the temperature anomaly index of each spatial region in the real-time temperature field based on spatial difference data;

[0027] Initial candidate regions are determined using a graph clustering algorithm based on the temperature anomaly index;

[0028] Based on the location data of historical disturbance areas and the current location of the friction head, the spatial position stability coefficient of each initial candidate area is calculated;

[0029] When the stability coefficient between adjacent candidate regions exceeds a preset threshold, the adjacent regions are merged into one candidate region.

[0030] When the stability coefficient does not exceed the preset threshold, the candidate regions remain independent;

[0031] Based on the regional merging results, the temperature difference magnitude and temperature change rate of each candidate region are calculated to determine the perturbation characteristics of each perturbation candidate region.

[0032] Furthermore, the calculation process for the spatial position stability coefficient includes:

[0033] The initial position deviation coefficient is determined based on the Euclidean distance between the center of the historical disturbance area and the center of the current candidate area;

[0034] Determine the position deviation weighting coefficient based on the current process status parameters;

[0035] Based on the movement trend of the center of the historical disturbance area, determine the historical trend stability coefficient;

[0036] The initial position deviation coefficient, the position deviation weight coefficient, and the historical trend stability coefficient are combined to calculate the spatial position stability coefficient.

[0037] Furthermore, the process of determining the location and intensity of real-time disturbance anomaly sources includes:

[0038] The initial disturbance intensity of the region is calculated based on the temperature difference data of each point within the candidate region.

[0039] Based on the initial disturbance intensity, calculate the location probability distribution of the real-time disturbance anomaly source corresponding to each location within the candidate region;

[0040] The location with the highest probability is selected as the location center of the real-time disturbance anomaly source based on the location probability distribution, and the average temperature difference in the adjacent area is calculated as the disturbance intensity based on this location.

[0041] Furthermore, the disturbance intensity calculation process includes:

[0042] The initial temperature difference value is calculated based on the difference between the real-time temperature data of the candidate area and the historical baseline temperature data.

[0043] S103.1.2 Determine the process state influence coefficient based on the current process state parameters, and use the process state influence coefficient to correct the initial temperature difference value;

[0044] The temperature fluctuation coefficient is determined based on the fluctuation frequency of temperature data within the candidate region.

[0045] The disturbance intensity of the candidate region is determined by fusing the corrected temperature difference value with the temperature fluctuation coefficient.

[0046] Furthermore, the process of determining the adjustment parameters for the corrected friction head state includes:

[0047] The initial friction head state adjustment parameters are calculated based on the location and intensity of the real-time disturbance anomaly source.

[0048] Calculate the trajectory correction factor based on the spatial relationship between the current friction head trajectory and the real-time disturbance anomaly source;

[0049] The initial friction head state adjustment parameters are adjusted based on the trajectory correction factor to determine the corrected friction head state adjustment parameters.

[0050] Secondly, the present invention provides a temperature field identification and control system for triboelectric solid-state additive manufacturing, implemented based on the aforementioned temperature field identification and control method for triboelectric solid-state additive manufacturing, comprising:

[0051] The data acquisition module is used to acquire real-time temperature field data and current process state parameters during the triboelectric solid-phase additive manufacturing process; the real-time temperature field data is subjected to joint decoupling processing in the spatial and temporal domains using a pre-built space-time temperature field coupling model to obtain spatial domain feature weights and temporal domain feature weights.

[0052] The difference processing module is used to determine the spatial difference data between the real-time temperature field and the reference temperature field by comparing the real-time temperature field data with the spatial domain feature weights and the time domain feature weights; and to determine the candidate regions for local abnormal disturbances and the disturbance features corresponding to each region based on the spatial difference data.

[0053] The parameter correction module is used to determine the location and intensity of the real-time disturbance anomaly source based on the disturbance characteristics of the candidate region and the current process state parameters; determine the initial friction head state adjustment parameters based on the location and intensity of the real-time disturbance anomaly source, and correct them based on the friction head trajectory to obtain the corrected friction head state adjustment parameters;

[0054] The update control module is used to adjust the feed trajectory of the friction head in real time according to the adjustment parameters of the corrected friction head state; and to dynamically update the space-time temperature field coupling model according to the updated real-time temperature field data for the next real-time temperature field data processing.

[0055] The technical effects and advantages provided by the present invention in the above technical solution are as follows:

[0056] This invention constructs a feature decoupling model based on spatial-temporal temperature field coupling, which enables real-time and accurate joint spatial and temporal analysis of temperature field data during the triboelectric solid-phase additive manufacturing process. This allows for precise identification of local temperature anomalies, significantly improving the accuracy of real-time temperature field analysis and the precision of anomaly location identification. It also avoids the problem in existing technologies where local anomalies cannot be accurately and promptly detected due to the limited monitoring methods.

[0057] This invention proposes a method for determining disturbance regions based on the fusion of spatial difference data and current process state parameters. This method can accurately and in real time determine the specific location and intensity characteristics of disturbance anomaly sources, enabling the friction head to respond accurately and quickly to local disturbance changes during the process. This effectively solves the problems of response lag and inaccurate control caused by the fixed threshold monitoring method in the prior art, and significantly improves the real-time response capability of temperature anomaly disturbance regions.

[0058] This invention establishes a dynamic feedback mechanism between the state parameters of the friction head and the disturbance characteristics, which enables real-time automatic adjustment of the feed trajectory and operating parameters of the friction head, and timely dynamic updating of the temperature field identification and control model, forming a complete real-time feedback control closed-loop mechanism. This effectively overcomes the problems of the single method of friction head state adjustment and difficulty in achieving real-time precise control in the prior art, and greatly improves the stability of the friction solid phase additive manufacturing process and the forming quality of the manufactured products. Attached Figure Description

[0059] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0060] Figure 1 This is a flowchart of a temperature field identification and control method for triboelectric solid-phase additive manufacturing according to the present invention;

[0061] Figure 2 This is a framework diagram of a temperature field identification and control system for triboelectric solid-phase additive manufacturing according to the present invention. Detailed Implementation

[0062] Exemplary embodiments will now be described more fully with reference to the accompanying drawings. However, these exemplary embodiments can be implemented in many forms and should not be construed as limited to the examples set forth herein; rather, they are provided to make the description of this application more complete and comprehensive, and to fully convey the concept of the exemplary embodiments to those skilled in the art. The drawings are merely illustrative illustrations of this application and are not necessarily drawn to scale. The same reference numerals in the drawings denote the same or similar parts, and therefore repeated descriptions of them will be omitted.

[0063] Furthermore, the described features, structures, or characteristics can be combined in any suitable manner in one or more exemplary embodiments. Numerous specific details are provided in the following description to give a full understanding of the exemplary embodiments disclosed in this application. However, those skilled in the art will recognize that the technical solutions disclosed in this application can be practiced with one or more specific details omitted, or other methods, components, steps, etc., can be employed. In other instances, well-known structures, methods, implementations, or operations are not shown or described in detail to avoid obscuring various aspects of the disclosure of this application.

[0064] Example 1

[0065] like Figure 1 As shown, this embodiment discloses a method for temperature field identification and control in triboelectric solid-state additive manufacturing, including:

[0066] S101: Obtain real-time temperature field data and current process state parameters during the triboelectric solid-phase additive manufacturing process; use a pre-built space-time temperature field coupling model to perform joint decoupling processing of the real-time temperature field data in the spatial and time domains to obtain spatial domain feature weights and time domain feature weights.

[0067] In this embodiment, real-time temperature field data refers to the spatial temperature information acquired in real time during the triboelectric solid-phase additive manufacturing process by a thermal imaging device or thermocouple array sensor installed on the additive manufacturing equipment. Specifically, the real-time temperature field data is a sequence of multi-point temperature measurement data taken at fixed time intervals (e.g., 10 times per second) during the manufacturing process. Each time data is collected, the additive region is divided into multiple spatial grids (e.g., the measurement region is divided into 10×10 spatial grids centered on the friction head's action area), and the real-time temperature value of each grid region is recorded, forming a spatially distributed temperature data array. For example, each sampling can yield data in the form of:

[0068] A two-dimensional real-time temperature data array, wherein T i,j This represents the real-time temperature value measured within the spatial grid in the i-th row and j-th column. The area of ​​each grid region is a predetermined specific size (e.g., 1cm × 1cm). With this clear spatial distribution, the real-time temperature field data can comprehensively and objectively reflect the real-time temperature of various local areas in the actual process, ensuring the accuracy of subsequent data processing.

[0069] The current process parameters involved in this embodiment specifically include the friction head feed speed, friction head rotation speed, pressing amount, and the physical property parameters of the processed material, specifically:

[0070] Friction head feed speed: The speed at which the friction head moves along the machining trajectory is recorded in real time, in mm / s, and is obtained by a linear displacement sensor installed on the feed device.

[0071] Friction head rotation speed: The angular velocity of the friction head rotation is recorded in real time, in rpm (revolutions per minute), and is measured in real time by a rotary encoder;

[0072] Indentation: The depth of pressure applied to the material surface by the friction head in real time (indentation depth), in mm, is obtained by real-time measurement using a piezoelectric or resistance strain gauge displacement sensor;

[0073] Physical property parameters of processed materials: Record the thermal physical parameters of the processed materials, such as thermal conductivity, thermal diffusivity, and specific heat capacity. These parameters are predetermined through material property databases or material performance manuals and are explicitly input into the control system before processing.

[0074] The current process status parameters are acquired in real time by the aforementioned sensors and transmitted synchronously to the data processing system in real time. Together with the real-time temperature field data, they form an accurate data foundation to ensure the accuracy of subsequent real-time temperature field abnormal disturbance identification and real-time adjustment of the friction head status.

[0075] In implementation, the joint decoupling process of the space-time temperature field coupling model includes:

[0076] A spatial-temporal feature tensor is constructed based on historical temperature field data, and an improved three-dimensional convolutional attention autoencoder is used to determine the initial spatial feature weights of each spatial region and the initial temporal feature weights of each time step.

[0077] In specific implementation, the space-time feature tensor is a three-dimensional data structure, specifically temperature data with a clear spatial distribution for each of several consecutive historical moments (e.g., 100 moments).

[0078] The improved 3D convolutional attention autoencoder model structure explicitly includes an encoder and a decoder:

[0079] Encoder: The input layer receives 3D feature tensor data; 3D convolutional layer (Conv3D): The convolution kernel size is 3×3×3, the stride is 1, the number of channels is explicitly defined as 16, 32, and 64, and the activation function is ReLU; Dual attention module: Attention weights are calculated in the spatial and temporal dimensions respectively; Max pooling layer: The pooling size is explicitly defined as 2×2×2;

[0080] Decoder: 3D deconvolutional layer (Conv3DTranspose): kernel size is 3×3×3, and the number of feature channels is explicitly defined as 64, 32, and 16 respectively; Output layer: outputs the initial spatial feature weights and initial temporal feature weights vectors respectively.

[0081] The temperature change rate of each spatial region is determined based on real-time temperature field data, and the initial spatial feature weights are adjusted using the temperature change rate to obtain the spatial domain feature weights.

[0082] Specifically, adjusting the initial spatial feature weights using the temperature change rate includes:

[0083] The temperature change sensitivity index for each spatial region is determined based on the rate of temperature change.

[0084] Specifically, based on the temperature difference (ΔT) between two adjacent time points for each spatial region, the temperature change rate of each spatial region is calculated as ΔT / Δt (where Δt is the sampling interval).

[0085] Based on a comparison of the magnitude of the temperature change rate in each spatial region with the preset normal fluctuation range, the temperature change sensitivity index for each spatial region is determined. The specific calculation formula is as follows:

[0086]

[0087] In the formula: SI i Let r be the sensitivity index for the i-th spatial region. ir is the actual calculated rate of temperature change. norm This represents the average rate of temperature change under historical normal conditions; a higher temperature change sensitivity index indicates a greater degree of temperature anomaly in the region.

[0088] Calculate the probability of temperature anomalies in a spatial region based on the temperature change sensitivity index;

[0089] Furthermore, the aforementioned sensitivity index SI is input into a probability transformation function (such as a softmax normalization function) to calculate the temperature anomaly probability for each spatial region, specifically:

[0090]

[0091] In the formula: P i Let N be the probability of an anomaly in the i-th spatial region, and N be the total number of regions. The temperature anomaly probability objectively reflects the likelihood of an anomaly occurring in each spatial region.

[0092] The initial spatial feature weights are adjusted by weighting the initial spatial feature weights based on the probability of temperature anomalies, and the spatial domain feature weights are determined accordingly.

[0093] Specifically, the initial spatial feature weights are adjusted based on the aforementioned anomaly probabilities, specifically employing a linear fusion of the initial spatial weights and the temperature anomaly probabilities:

[0094]

[0095] In the formula: W s,i The adjusted spatial domain feature weights, The initial spatial feature weights are used as the basis for adjusting the spatial domain feature weights to more accurately reflect the anomaly sensitivity of the real-time temperature field spatial region.

[0096] The temperature change rate at each moment is determined based on real-time temperature field data, and the initial time feature weights are adjusted using the temperature change rate to obtain the time domain feature weights.

[0097] Specifically, adjusting the initial time feature weights using the rate of temperature change includes:

[0098] Determine the temperature change sensitivity index at each moment based on the rate of temperature change:

[0099]

[0100] In the formula: V t T is the temperature change sensitivity index. i (t) represents the temperature value of the i-th spatial region at time t, T i (t-1) represents the temperature value at time t-1, and M represents the total number of spatial regions;

[0101] Determine the temperature anomaly trend coefficient based on the temperature change sensitivity index:

[0102] C t =αV t +(1-α)C t-1 ;

[0103] In the formula: C t The coefficient represents the temperature anomaly trend, and α is the smoothing factor, typically set to 0.3.

[0104] The initial time-domain feature weights are adjusted based on the temperature anomaly trend coefficient to determine the time-domain feature weights:

[0105]

[0106] In the formula: W t,k The adjusted time-domain feature weights, These are the initial time feature weights.

[0107] S102: Based on the real-time temperature field data and the spatial domain feature weights and time domain feature weights, determine the spatial difference data between the real-time temperature field and the reference temperature field; based on the spatial difference data, determine the candidate regions for local abnormal disturbances and the disturbance characteristics corresponding to each region;

[0108] Specifically, the method for determining the spatial difference data between the real-time temperature field and the reference temperature field includes:

[0109] The weighted difference between the temperature data at each spatial location in the real-time temperature field and the corresponding temperature data in the reference temperature field is calculated, and the formula is expressed as follows:

[0110]

[0111] In the formula, ΔT i,j (t) represents the spatial difference data of the i,j position at time t; W s,i,j Spatial domain feature weights; W t (t) represents the time-domain feature weights; This refers to the real-time temperature field. The reference temperature field temperature;

[0112] In implementation, the process of determining candidate regions for local anomalous disturbances and the corresponding disturbance characteristics of each region includes:

[0113] Calculate the temperature anomaly index of each spatial region in the real-time temperature field based on spatial difference data;

[0114]

[0115] In the formula: AI i,j(t) represents the temperature anomaly index; Let σ be the mean of all spatial difference data at time t. ΔT (t) represents the standard deviation; the above index measures the degree of anomaly of each spatial region relative to the overall temperature field.

[0116] The initial candidate regions are determined using a graph clustering algorithm based on the temperature anomaly index. The specific implementation method is as follows:

[0117] Using each spatial location as a node, the anomaly index as the node weight, and the spatial adjacency relationship between nodes to determine the edges of the graph, an explicit spectral clustering algorithm is used to divide the nodes and form several initial candidate regions.

[0118] For example, a temperature anomaly index greater than a set threshold (e.g., |AI) can be used. i,j The spatial nodes (t)>2) are used as core nodes and divided into initial candidate regions by spectral clustering method;

[0119] Based on the location data of historical disturbance areas and the current location of the friction head, the spatial position stability coefficient of each initial candidate area is calculated;

[0120] Specifically, the calculation process of the spatial position stability coefficient includes:

[0121] The initial position deviation coefficient is determined based on the Euclidean distance between the center of the historical disturbance area and the center of the current candidate area;

[0122]

[0123] In the formula: (x h ,y h (x) represents the center of the historical disturbance area. c ,y c () represents the center location of the current candidate region;

[0124] Determine the position deviation weighting coefficient based on the current process status parameters;

[0125]

[0126] In the formula: v f d represents the feed rate of the friction head, and d represents the compression amount. With d max The value of k is the historical maximum value of the corresponding process parameter. v With k d These are the weighting coefficients (for example, set to 0.6 and 0.4 respectively);

[0127] Based on the movement trend of the center of the historical disturbance area, determine the historical trend stability coefficient;

[0128] Specifically, firstly, the trend of historical center location change in the disturbed region is fitted using linear regression to obtain the slope s of the location change; explicitly, the historical trend stability coefficient HT is calculated as follows:

[0129] HT = e -|s| ;

[0130] The initial position deviation coefficient, the position deviation weight coefficient, and the historical trend stability coefficient are combined to calculate the spatial position stability coefficient.

[0131]

[0132] For example, if the initial position deviation coefficient PD = 2.5, the position deviation weighting coefficient PW = 0.8, and the historical trend stability coefficient HT = 0.9, then the final stability coefficient is:

[0133]

[0134] When the stability coefficient between adjacent candidate regions exceeds a preset threshold, the adjacent regions are merged into one candidate region.

[0135] When the stability coefficient does not exceed the preset threshold, the candidate regions remain independent;

[0136] Based on the regional merging results, the temperature difference magnitude and temperature change rate of each candidate region are calculated to determine the perturbation characteristics of each perturbation candidate region.

[0137] Among them, the temperature difference amplitude is specifically defined as the spatial difference data ΔT within the region. i,j The average value of (t):

[0138]

[0139] Where: N r R represents the number of nodes in the region, and R represents all spatial points in the candidate region.

[0140] Specifically, the rate of temperature change is defined as the change in the magnitude of the average temperature difference within the region between the current and previous time points.

[0141]

[0142] S103: Determine the location and intensity of the real-time disturbance anomaly source based on the disturbance characteristics of the candidate region and the current process state parameters; determine the initial friction head state adjustment parameters based on the location and intensity of the real-time disturbance anomaly source, and correct them based on the friction head trajectory to obtain the corrected friction head state adjustment parameters;

[0143] In practice, the process of determining the location and intensity of real-time disturbance anomaly sources includes:

[0144] The initial disturbance intensity of the region is calculated based on the temperature difference data of each point within the candidate region.

[0145] Specifically, the disturbance intensity calculation process includes:

[0146] The initial temperature difference value is calculated based on the difference between the real-time temperature data of the candidate area and the historical baseline temperature data.

[0147]

[0148] The process state influence coefficient is determined based on the current process state parameters, and the initial temperature difference value is corrected using the process state influence coefficient.

[0149] Specifically, firstly, the feed speed v of the friction head... f Rotational speed v r The normalized values ​​of the compression amount d and the pressure amount d are calculated separately, and then the influence coefficient K is determined by linear weighting. p :

[0150]

[0151] In the formula, w f ,w r ,w d The weighting coefficients for feed rate, rotational speed and compression amount are specified respectively (e.g., set to 0.3, 0.4 and 0.3 respectively);

[0152] The initial temperature difference value is further corrected using the process condition influence coefficient to determine the corrected temperature difference value DT. corr The formula is:

[0153] DT corr =DT init ×(1+K p )

[0154] The temperature fluctuation coefficient is determined based on the fluctuation frequency of temperature data within the candidate region.

[0155] Specifically, Fourier spectrum analysis is performed on the temperature data at each spatial location within the candidate region to determine the dominant fluctuation frequency f of the regional temperature data. dom And explicitly calculate the volatility coefficient:

[0156]

[0157] The corrected temperature difference value is fused with the temperature fluctuation coefficient to determine the disturbance intensity of the candidate region;

[0158] DS=DT corr ×Kf ;

[0159] In the formula: DS is the disturbance intensity;

[0160] Based on the initial disturbance intensity, calculate the location probability distribution of the real-time disturbance anomaly source corresponding to each location within the candidate region;

[0161]

[0162] Where: ΔT i,j (t) represents the spatial difference data for each spatial location within the region, and the sum of probabilities within the region is 1;

[0163] Based on the location probability distribution, the location with the highest probability is selected as the location center of the real-time disturbance anomaly source, and the average temperature difference in the adjacent area is calculated as the disturbance intensity with this location as the center.

[0164] Specifically, based on the above location probability distribution, the spatial location with the highest probability value is selected as the location center of the real-time disturbance anomaly source. Then, spatial points within the adjacent area (such as a 3×3 spatial region centered on this point) are selected, and the average temperature difference of this region is calculated to determine the final disturbance intensity.

[0165]

[0166] Where: N c Let C be the number of points in the neighboring region, and C be the set of coordinates of all points in the neighboring region.

[0167] In practice, the process of determining the adjustment parameters for the corrected friction head state includes:

[0168] The initial friction head state adjustment parameters are calculated based on the location and intensity of the real-time disturbance anomaly source.

[0169] Specifically, based on the location center coordinates (x, y) of the real-time disturbance anomaly source. s ,y s ) and final disturbance intensity DS final Clearly calculate the initial friction head state adjustment parameters, including the adjustment amount of the initial feed speed of the friction head. Rotation speed adjustment amount and the amount of pressure adjustment Δd init :

[0170]

[0171] In the formula: k1, k2, k3 are coefficients that match the control characteristics of the equipment;

[0172] Calculate the trajectory correction factor based on the spatial relationship between the current friction head trajectory and the real-time disturbance anomaly source;

[0173]

[0174] In the formula: (x t ,y t (x) represents the coordinates of the predetermined target position on the current friction head trajectory; s ,y s () represents the center coordinates of the real-time disturbance anomaly source location;

[0175] The initial friction head state adjustment parameters are adjusted according to the trajectory correction factor to determine the corrected friction head state adjustment parameters;

[0176]

[0177] For example, if the initial friction head feed speed adjustment is If the trajectory correction factor TF = 0.8, then the corrected feed rate adjustment is:

[0178] S104: Adjust the feed trajectory of the friction head in real time according to the adjusted parameters of the friction head state; and dynamically update the space-time temperature field coupling model according to the updated real-time temperature field data for the next real-time temperature field data processing.

[0179] In specific implementation, the real-time adjustment of the friction head feed trajectory includes:

[0180] Based on the corrected adjustment amounts of the friction head feed speed, rotation speed, and compression, a clear friction head state control command is generated and sent to the friction head motion control unit in real time to dynamically correct the current machining trajectory of the friction head.

[0181] For example, if the current friction head feed speed is 5.00 mm / s, the rotation speed is 600 rpm, the indentation is 0.20 mm, and the correction parameters are as follows:

[0182] Feed rate adjustment amount:

[0183] Rotation speed adjustment amount:

[0184] Adjustment amount of compression: Δd corr =0.05mm;

[0185] The real-time adjusted feed rate, rotational speed, and compression amount of the friction head are determined as follows:

[0186]

[0187] Subsequently, the motion control unit adjusts the friction head trajectory in real time according to the above-mentioned real-time adjusted parameters in a clear closed-loop feedback control method, so that the friction head trajectory tends to the center position of the real-time disturbance abnormal source area, thereby reducing the temperature abnormal disturbance area and improving manufacturing quality.

[0188] Specifically, based on the real-time temperature field data after adjusting the above-mentioned friction head feed trajectory, the space-time temperature field coupling model is dynamically updated, and the specific implementation method includes:

[0189] Real-time acquisition of spatial-temporal temperature field data after feed trajectory adjustment to form a new real-time temperature field data sample set, specifically including the updated real-time temperature field data matrix T. new (t);

[0190] An incremental update method is used to dynamically update the space-time temperature field coupling model. The specific implementation process is as follows:

[0191] (1) The newly acquired real-time temperature field data sample T new (t) is merged with historical data samples to form an extended data sample set;

[0192] (2) Extract the most recent fixed-window-length data (e.g., the most recent 100 time-time data) from the expanded data sample set to form a new training sample set for dynamic training of the space-time temperature field coupling model.

[0193] (3) Using the new training sample set as input, the original model parameters are incrementally updated using an adaptive dynamic learning algorithm. Specifically, the adaptive dynamic learning algorithm is defined as Online Adaptive Stochastic Gradient Descent (Online-SGD), and the parameter update formula is:

[0194]

[0195] In the formula: θ t θ represents the current model parameters. t+1 For the updated model parameters, η t For adaptive learning rate (e.g., initially 0.001), The gradient of the loss function on the current real-time data sample;

[0196] Specifically, the loss function L is defined as the mean squared error (MSE), and the formula is explicitly expressed as:

[0197]

[0198] In the formula: T new (t) represents the newly input real-time temperature field data. The reconstructed data is the output of the model, where N is the total number of data samples.

[0199] After the model is dynamically updated, the spatial domain feature weights and time domain feature weights are updated in real time to achieve accurate processing of the next real-time temperature field data.

[0200] For example, after a model dynamic update, if the prediction accuracy of the new real-time temperature field data is significantly improved (e.g., the prediction error decreases by more than 20%), the new parameters will replace the old parameters and be used for subsequent real-time temperature field data analysis.

[0201] Example 2

[0202] like Figure 2 As shown in the example, the parts not detailed in this embodiment are as shown in Example 1. This embodiment discloses a temperature field identification and control system for triboelectric solid-state additive manufacturing, including:

[0203] The data acquisition module 201 is used to acquire real-time temperature field data and current process state parameters during the triboelectric solid-phase additive manufacturing process; and to perform joint decoupling processing of the real-time temperature field data in the spatial domain and time domain using a pre-constructed space-time temperature field coupling model to obtain spatial domain feature weights and time domain feature weights.

[0204] The difference processing module 202 is used to determine the spatial difference data between the real-time temperature field and the reference temperature field by comparing the real-time temperature field data with the spatial domain feature weights and the time domain feature weights; and to determine the candidate regions for local abnormal disturbances and the disturbance features corresponding to each region based on the spatial difference data.

[0205] The parameter correction module 203 is used to determine the location and intensity of the real-time disturbance anomaly source based on the disturbance characteristics of the candidate region and the current process state parameters; determine the initial friction head state adjustment parameters based on the location and intensity of the real-time disturbance anomaly source, and correct them based on the friction head trajectory to obtain the corrected friction head state adjustment parameters;

[0206] The update control module 204 is used to adjust the feed trajectory of the friction head in real time according to the adjustment parameters of the corrected friction head state; and to dynamically update the space-time temperature field coupling model according to the updated real-time temperature field data for the next real-time temperature field data processing.

[0207] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters, weights, and thresholds in the formulas are set by those skilled in the art according to the actual situation.

[0208] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, all or part of the processes or functions described in the embodiments of the present invention are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via a wired or wireless network. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more sets of available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium. A semiconductor medium can be a solid-state drive.

[0209] The foregoing has only described certain exemplary embodiments of the present invention by way of illustration. Undoubtedly, those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the foregoing drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

Claims

1. A method for identifying and controlling the temperature field in triboelectric solid-phase additive manufacturing, characterized in that, include: S101: Acquire real-time temperature field data and current process parameters during the triboelectric solid-phase additive manufacturing process; A pre-built spatial-temporal temperature field coupling model is used to perform joint decoupling processing of real-time temperature field data in the spatial and temporal domains to obtain spatial domain feature weights and temporal domain feature weights. S102: Determine the spatial difference data between the real-time temperature field and the reference temperature field based on the real-time temperature field data and the spatial domain feature weights and time domain feature weights; Based on spatial difference data, candidate regions for local anomalous disturbances and the corresponding disturbance characteristics of each region are determined; S103: Determine the location and intensity of the real-time disturbance anomaly source based on the disturbance characteristics of the candidate region and the current process status parameters; The initial friction head state adjustment parameters are determined based on the location and intensity of the real-time disturbance anomaly source, and then corrected based on the friction head trajectory to obtain the corrected friction head state adjustment parameters. S104: Adjust the feed trajectory of the friction head in real time according to the adjusted parameters of the friction head state; The space-time temperature field coupling model is dynamically updated based on the updated real-time temperature field data for use in the next real-time temperature field data processing.

2. The method for temperature field identification and control in triboelectric solid-state additive manufacturing according to claim 1, characterized in that, The joint spatial and temporal decoupling processing of real-time temperature field data includes: A spatial-temporal feature tensor is constructed based on historical temperature field data, and an improved three-dimensional convolutional attention autoencoder is used to determine the initial spatial feature weights of each spatial region and the initial temporal feature weights of each time step. The temperature change rate of each spatial region is determined based on real-time temperature field data, and the initial spatial feature weights are adjusted using the temperature change rate to obtain the spatial domain feature weights. The temperature change rate at each moment is determined based on real-time temperature field data, and the initial time feature weights are adjusted using the temperature change rate to obtain the time domain feature weights.

3. The method for temperature field identification and control in triboelectric solid-phase additive manufacturing according to claim 2, characterized in that, The adjustment of the initial spatial feature weights using the temperature change rate includes: The temperature change sensitivity index for each spatial region is determined based on the rate of temperature change. Calculate the probability of temperature anomalies in a spatial region based on the temperature change sensitivity index; The initial spatial feature weights are adjusted by weighting the initial spatial feature weights based on the probability of temperature anomalies, and the spatial domain feature weights are determined.

4. The method for temperature field identification and control in triboelectric solid-state additive manufacturing according to claim 3, characterized in that, The adjustment of the initial time feature weights using the rate of temperature change includes: The temperature change sensitivity index is determined based on the rate of temperature change. Determine the temperature anomaly trend coefficient based on the temperature change sensitivity index; The initial time feature weights are adjusted by weighting the initial time feature weights based on the temperature anomaly trend coefficient, and the time domain feature weights are determined.

5. The method for temperature field identification and control in triboelectric solid-phase additive manufacturing according to claim 4, characterized in that, The process of determining candidate regions for local anomalous disturbances and the corresponding disturbance characteristics of each region includes: Calculate the temperature anomaly index of each spatial region in the real-time temperature field based on spatial difference data; Initial candidate regions are determined using a graph clustering algorithm based on the temperature anomaly index; Based on the location data of historical disturbance areas and the current location of the friction head, the spatial position stability coefficient of each initial candidate area is calculated; When the stability coefficient between adjacent candidate regions exceeds a preset threshold, the adjacent regions are merged into one candidate region. When the stability coefficient does not exceed the preset threshold, the candidate regions remain independent; Based on the regional merging results, the temperature difference magnitude and temperature change rate of each candidate region are calculated to determine the perturbation characteristics of each perturbation candidate region.

6. The method for temperature field identification and control in triboelectric solid-state additive manufacturing according to claim 5, characterized in that, The calculation process for the spatial position stability coefficient includes: The initial position deviation coefficient is determined based on the Euclidean distance between the center of the historical disturbance area and the center of the current candidate area; Determine the position deviation weighting coefficient based on the current process status parameters; Based on the movement trend of the center of the historical disturbance area, determine the historical trend stability coefficient; The initial position deviation coefficient, the position deviation weight coefficient, and the historical trend stability coefficient are combined to calculate the spatial position stability coefficient.

7. The method for temperature field identification and control in triboelectric solid-phase additive manufacturing according to claim 6, characterized in that, The process of determining the location and intensity of real-time disturbance anomaly sources includes: The initial disturbance intensity of the region is calculated based on the temperature difference data of each point within the candidate region. Based on the initial disturbance intensity, calculate the location probability distribution of the real-time disturbance anomaly source corresponding to each location within the candidate region; The location with the highest probability is selected as the location center of the real-time disturbance anomaly source based on the location probability distribution, and the average temperature difference in the adjacent area is calculated as the disturbance intensity based on this location.

8. The method for temperature field identification and control in triboelectric solid-state additive manufacturing according to claim 7, characterized in that, The disturbance intensity calculation process includes: The initial temperature difference value is calculated based on the difference between the real-time temperature data of the candidate area and the historical baseline temperature data. The process state influence coefficient is determined based on the current process state parameters, and the initial temperature difference value is corrected using the process state influence coefficient. The temperature fluctuation coefficient is determined based on the fluctuation frequency of temperature data within the candidate region. The disturbance intensity of the candidate region is determined by fusing the corrected temperature difference value with the temperature fluctuation coefficient.

9. The method for temperature field identification and control in triboelectric solid-state additive manufacturing according to claim 8, characterized in that, The process of determining the adjustment parameters for the corrected friction head state includes: The initial friction head state adjustment parameters are calculated based on the location and intensity of the real-time disturbance anomaly source. Calculate the trajectory correction factor based on the spatial relationship between the current friction head trajectory and the real-time disturbance anomaly source; The initial friction head state adjustment parameters are adjusted based on the trajectory correction factor to determine the corrected friction head state adjustment parameters.

10. A temperature field identification and control system for triboelectric solid-state additive manufacturing, implemented based on the temperature field identification and control method for triboelectric solid-state additive manufacturing according to any one of claims 1-9, characterized in that, include: The data acquisition module is used to acquire real-time temperature field data and current process parameters during the triboelectric solid-phase additive manufacturing process. A pre-built spatial-temporal temperature field coupling model is used to perform joint decoupling processing of real-time temperature field data in the spatial and temporal domains to obtain spatial domain feature weights and temporal domain feature weights. The difference processing module is used to determine the spatial difference data between the real-time temperature field and the reference temperature field by comparing the real-time temperature field data with the spatial domain feature weights and the time domain feature weights. Based on spatial difference data, candidate regions for local anomalous disturbances and the corresponding disturbance characteristics of each region are determined; The parameter correction module is used to determine the location and intensity of real-time disturbance anomaly sources based on the disturbance characteristics of the candidate region and the current process state parameters; The initial friction head state adjustment parameters are determined based on the location and intensity of the real-time disturbance anomaly source, and then corrected based on the friction head trajectory to obtain the corrected stirring head state adjustment parameters. The control module is updated to adjust the feed trajectory of the friction head in real time according to the adjustment parameters of the corrected friction head state. The space-time temperature field coupling model is dynamically updated based on the updated real-time temperature field data for use in the next real-time temperature field data processing.

Images