Multi-nozzle collaborative control method and system for 3D printing

Abstract

The present invention belongs to the field of 3D printing technology, and in particular, relates to a multi-nozzle collaborative control method and system for 3D printing. The method pre-constructs a three-dimensional parameter model of a to-be-printed part and refines the stratification and regional division, combines the process attributes of each region with the characteristics of the nozzle, and intelligently distributes printing tasks to multiple nozzles. Secondly, based on optimized printing path planning, multi-nozzle collaborative operation is achieved. During the printing process, non-destructive testing technology is used to monitor and predict printing index deviations in real time. Through the built-in error trend prediction model, nozzle parameters are adjusted in advance to ensure that the printing indicators meet the preset process requirements in real time. The multi-nozzle collaborative control method of the present invention significantly improves the accuracy and efficiency of 3D printing of complex structural parts.

Description

Technical Field

[0001] The present invention belongs to the technical field of 3D printing, and in particular relates to a multi-nozzle collaborative control method and system for 3D printing. Background Art

[0002] With the advancement of technology, multi-nozzle 3D printing technology has gradually become a research hotspot in the field of jewelry manufacturing. This technology usually uses multiple nozzles to work together, and each nozzle can be loaded with different materials, such as precious metal powder, resin or other special materials. This not only increases the printing speed, but also enables integrated printing of multiple materials, providing more possibilities for jewelry design. Although multi-nozzle 3D printing technology has shown great potential in jewelry manufacturing, the existing technology still has some limitations and challenges. For example, the existing multi-nozzle 3D printing system is not mature enough in terms of coordinated control between nozzles, especially when dealing with complex structures. How to efficiently assign tasks to each nozzle is still a challenge;

[0003] For example, a Chinese patent with publication number CN115771259A discloses a 3D printing online coloring collaborative control method and system, including: initializing the position of the print head and the coloring nozzle of the 3D printer; when the print head completes the printing of the preset M layers, determining the coloring position of the coloring nozzle, and controlling the coloring nozzle to move to the coloring position; when the coloring nozzle moves to the coloring position, triggering the coloring nozzle to start working; determining a first distance and a second distance based on a first infrared probe arranged on the print head and a second infrared probe arranged on the coloring nozzle; when the first distance and the second distance are both greater than or equal to a first preset distance threshold, controlling the print head and the coloring nozzle to continue to work simultaneously; when the print head completes the printing of the current printing layer, or the coloring nozzle completes the coloring of the current coloring layer and the layer number difference between the coloring layer and the printing layer is greater than or equal to the preset layer number threshold, printing the next layer or coloring the next layer.

[0004] The above prior art has the following problems: When dealing with complex structures, how to perform task allocation and collaborative control after task allocation is still a challenge. In order to solve the above problems, the present invention provides a multi-nozzle collaborative control method and system for 3D printing. Summary of the invention

[0005] In view of the shortcomings of the prior art, the present invention proposes a multi-nozzle collaborative control method and system for 3D printing. The method pre-constructs a three-dimensional parameter model of the to-be-printed part and refines the stratification and regional division, combines the process attributes of each area with the nozzle characteristics, and intelligently allocates printing tasks to multiple nozzles. Secondly, based on the optimized printing path planning, the collaborative operation of multiple nozzles is realized. During the printing process, non-destructive testing technology is used to monitor and predict the printing index deviation in real time. Through the built-in error trend prediction model, the nozzle parameters are adjusted in advance to ensure that the printing index meets the preset process requirements in real time. The multi-nozzle collaborative control method of the present invention significantly improves the accuracy and efficiency of 3D printing of complex structural parts.

[0006] To achieve the above object, the present invention provides the following technical solutions:

[0007] The multi-nozzle collaborative control method for 3D printing includes:

[0008] S1. Obtain processing parameters of the workpiece to be printed, construct a three-dimensional parameter model of the workpiece to be printed by using three-dimensional software, and pre-layer and divide the constructed three-dimensional parameter model into regions, and obtain processing sub-parameters of the corresponding regions after the region division;

[0009] S2. According to the obtained processing sub-parameters of the corresponding area and the nozzle properties, the printing task types of the corresponding divided areas and the regional task execution list corresponding to the nozzle are obtained through a classification matching algorithm;

[0010] S3, initialize the multi-nozzle position, and generate the multi-nozzle segmented area path through the path generation algorithm according to the processing technology sub-parameters corresponding to each area, the printing task type and the area task execution list corresponding to the nozzle;

[0011] S4. Build a multi-agent collaborative control model through reinforcement learning, and build the multi-nozzle collaborative control strategy into the multi-nozzle segmented area path to control the dynamic 3D printing state of the multi-nozzle corresponding path in real time;

[0012] S5. Use non-destructive testing sensors to monitor the 3D printing process of each nozzle in real time, obtain the printing index data of the workpiece to be printed, obtain the printing index deviation according to the obtained printing index data and the processing parameters of the workpiece to be printed, and input the obtained printing index deviation into the built-in error trend prediction model to obtain the printing index prediction deviation of each nozzle at the next moment;

[0013] S6, feeding back the predicted deviation of the printing index of each nozzle at the next moment to the multi-agent collaborative control model, and making pre-adjustments to each nozzle during the printing process, so that the printing index data of the to-be-printed part is equal to the processing parameters of the to-be-printed part in real time;

[0014] The step of obtaining the processing sub-parameters of the corresponding regions after the region division in S1 includes:

[0015] S101, setting the number of pre-layers M and the layer thickness, dividing the three-dimensional parameter model into layers according to the preset number of layers M and the layer thickness through the software, obtaining a single printing layer, analyzing the geometric parameters and the required printing material parameters of the divided single printing layer through the 3D software, and obtaining a regional attribute parameter set and a processing technology parameter set for each layer after the division;

[0016] S102, inputting each acquired single-layer region attribute parameter set and processing technology parameter set into the grid algorithm to obtain each single-layer divided region sequence and the corresponding divided region processing technology sub-parameter set;

[0017] S103, setting a printing order of a single-layer printing layer, and setting a starting printing area in each obtained single-layer divided area sequence, and sorting the single-layer divided area sequence in ascending order according to the distance value between the center point of each divided area and the center point of the starting printing area position, obtaining a printing area sequence sequence, and marking the processing technology sub-parameter set of the corresponding area under each divided area in the obtained printing area sequence sequence;

[0018] S104, repeat S102 and S103 to obtain the M-layer division area space of the three-dimensional parameter model [ d mn ] M × N and divide the regional processing technology sub-parameter space, where represents the nth partition area in the mth layer, and N represents the total number of partition areas in a single layer;

[0019] Specifically, the specific steps of S2 include:

[0020] S201, obtain S104 [ d mn ] M × N And the corresponding partitioned regional processing technology sub-parameter space is input into the pre-trained regional task classification model to obtain The 3D printing task type space corresponding to the divided area [ d nm p ] M × N × P , the number of task types and the processing technology sub-parameter space corresponding to the task type [ d nm pk ] M × N × K ,in, represents the pth 3D printing task type in the nth partition area in the mth layer, represents the kth processing sub-parameter corresponding to the pth 3D printing task type in the nth divided area in the mth layer, K represents the total number of processing sub-parameters corresponding to the pth 3D printing task type; P represents the total number of task types in the nth divided area in the mth layer;

[0021] S202: Obtain nozzle attribute parameters, and use the obtained nozzle attribute parameters to construct a nozzle parameter space [ u j q ] J × Q ,in, represents the qth attribute parameter of the jth nozzle, J represents the total number of nozzles, and Q represents the total number of attribute parameters corresponding to each nozzle;

[0022] S203, the acquired 3D printing task type space [ d nm p ] M × N × P Corresponding processing technology sub-parameter space [ d nm pk ] M × N × K and nozzle parameter space [ u j q ] J × Q , input into the matching algorithm to obtain the jth nozzle and The matching degree of task types ,in Indicates the matching degree between the pth task type and the jth sprinkler in the nth partition area in the mth layer;

[0023] S204: Setting matching threshold ,when Greater than Then the pth 3D printing task type in the nth partition area in the mth layer is assigned to the jth nozzle, otherwise it is not assigned to the jth nozzle;

[0024] S205, repeat S204 process, from [ d nm p ] M × N × P Extracting satisfaction from Task type, construct the printing task set of the j-th nozzle [ b nm v ] V , It indicates that the j-th print head corresponds to the v-th print task in the print task set, and the v-th print task is located in the n-th partition area in the m-th layer, and V indicates the total number of tasks in the print task set corresponding to the j-th print head;

[0025] The specific steps of S2 also include:

[0026] S206, sorting the tasks in the print task set of the j-th nozzle according to the print area sequence obtained in S103, and making a one-to-one correspondence between the sorted tasks and the area numbers in the print area sequence, to obtain a print task sequence set of the j-th nozzle;

[0027] S207, repeat the process of S206 to obtain the printing task sequence space corresponding to J nozzles;

[0028] S208, creating a print task list, and embedding the obtained print task sequence space corresponding to the J nozzles and the processing technology sub-parameters corresponding to each task in the print task sequence space into the print task list, and obtaining a regional task execution list corresponding to the J nozzles.

[0029] Specifically, the steps of acquiring the multi-nozzle segmented area path include:

[0030] S301, inputting the printing task sequence set and corresponding processing technology sub-parameters of the j-th nozzle in the regional task execution list obtained in S208 into the path generation algorithm built into the 3D software to generate a task execution path corresponding to the j-th nozzle;

[0031] S302: insert V task start execution time points into the task execution path corresponding to the jth nozzle according to the print task sequence set corresponding to the jth nozzle and the corresponding processing technology sub-parameters. , the specific start time of task execution of the vth task is:

[0032]

[0033] in, Indicates the start time of the first task execution of the j-th nozzle, Indicates the starting execution time point of the v-1th task corresponding to the jth nozzle, Indicates the execution time length of the v-1th task of the jth sprinkler; Indicates the starting waiting time for the jth nozzle to perform the vth task;

[0034] S303, collect the historical start waiting time sequence and the corresponding printing index deviation of the j-th nozzle when executing the task in the printing task sequence set, use the historical start waiting time sequence and the corresponding printing index deviation to train the waiting time deviation prediction model constructed by the support vector machine, and obtain the starting waiting time deviation of the v-th task of the j-th nozzle ;

[0035] S304, feeding back the calculated starting waiting time deviation of the vth task printing task of the jth nozzle to the task starting execution time point of the vth task, adjusting the starting execution time point of the vth task execution corresponding to the jth nozzle, and obtaining the adjusted task starting execution time point of the vth task, specifically:

[0036]

[0037] S305, repeating the processes of S303 and S304 to adjust the task start execution time points of the V tasks corresponding to the j-th nozzle, and using the adjusted task start execution time points to segment the generated task execution path, to obtain the segmented area path corresponding to the j-th nozzle;

[0038] S306. Repeat the process of S301-S305 to obtain the segmented regional paths corresponding to the J nozzles, and save the segmented regional paths corresponding to the J nozzles into the regional task execution list to obtain the regional task path execution list.

[0039] Specifically, the multi-agent collaborative control model includes J single control sub-models and one collaborative control sub-model. The J single control sub-models correspond to the J sprinklers one by one. The specific steps of constructing the single control sub-model include:

[0040] S401, construct the current input state according to the segmented regional path of the j-th nozzle, the printing task sequence set and the processing sub-parameters corresponding to the printing task saved in the regional task path execution list of S306:

[0041] ,

[0042] in, Indicates the position of the j-th nozzle in the divided area at the current time t, Indicates the progress of the task execution in the printing task sequence set corresponding to the j-th nozzle at the current time t. It indicates the moving speed of the jth nozzle when performing the vth task at the current time t. It indicates the printing speed of the jth nozzle when executing the vth task at the current time t. It indicates the thickness corresponding to the task v performed by the j-th nozzle at the current time t. Indicates the starting time point corresponding to the jth nozzle executing the vth task, Indicates the printing index deviation corresponding to the j-th print head executing the v-th task at the current time t;

[0043] S402, setting the current moment execution trigger information based on the set current moment input state, specifically:

[0044] when The first execution action information is triggered when the corresponding execution adjustment and Actions, represents the processing thickness threshold corresponding to the vth task performed by the jth nozzle;

[0045] when The second execution action information is triggered when a single or combined execution adjustment is performed. , and Actions;

[0046] When the order of executing tasks corresponding to the j-th print head does not satisfy the order in the printing task sequence set, the third execution action information is triggered, that is, the action of adjusting the task execution sequence is executed;

[0047] S403: Construct the current execution action according to the current execution trigger information ,in, Indicates the action that triggers the first execution action information to be executed. Indicates the action that triggers the second execution action information to be executed. Indicates the action executed by triggering the third execution action information;

[0048] S404. Set the reward function according to the action executed at the current moment of the construction:

[0049]

[0050] in, It indicates the thickness error corresponding to the j-th nozzle performing the v-th task at the current time t, represents the reward error corresponding to the o-th adjustment of the task execution order, O represents the total number of times the j-th nozzle completes the task adjustment execution order in the corresponding printing task sequence set, is the gain coefficient corresponding to the reward function.

[0051] Specifically, the specific steps of the collaborative control sub-model work include:

[0052] S405, real-time monitoring to obtain the real-time processing parameters of J nozzles under the control of J single control sub-models and the movement parameters of the nozzles when performing tasks, and input the obtained movement parameters of the nozzles when performing tasks into the collision prediction model constructed by the support vector machine to obtain the probability sequence of the collision between the j-th nozzle and the g-th nozzle among the J nozzles at the current moment;

[0053] S406, setting a collision probability threshold, comparing the obtained collision probability sequence at the current moment with the collision probability threshold, and obtaining the nozzles and the corresponding collision probabilities corresponding to the collision probability at the current moment being greater than the collision probability threshold;

[0054] S407, feeding back the nozzles corresponding to the current moment collision probability greater than the collision probability threshold and the corresponding collision probability to the cooperative control sub-model constructed by the control algorithm, generating control signals for the nozzles corresponding to the current moment collision probability greater than the collision probability threshold, and transmitting the generated control signals to the single control sub-model of the corresponding nozzles to control the nozzles to perform collision avoidance actions;

[0055] S408, real-time monitoring of the real-time layer thickness parameters of the m-th single-layer printing layer completed by the nozzle, and comparing the real-time layer thickness parameters of the m-th single-layer printing layer with the layer thickness parameters in the processing parameter set of each layer after division obtained in S101. If they are not equal, the layer thickness error calculated by the real-time layer thickness parameters and the layer thickness parameters in the processing parameter set is fed back to the collaborative control sub-model to generate a collaborative control adjustment signal;

[0056] S409: Feedback the generated collaborative control adjustment signal to the single control sub-model corresponding to all nozzles executing the m-th single-layer printing task, triggering the model to perform the adjustment , or The action makes the real-time layer thickness parameter equal to the layer thickness parameter in the machining process parameter set.

[0057] The multi-nozzle collaborative control system for 3D printing includes: a three-dimensional module, a task allocation module, a printing control module and a monitoring feedback module;

[0058] The three-dimensional module includes a three-dimensional construction unit and a region division unit; the region division unit is used to pre-layer the three-dimensional parameter model constructed by the three-dimensional construction unit, and use a grid algorithm to perform virtual region division and annotation on each layer after pre-layering, and obtain the corresponding regional processing technology sub-parameters after regional division;

[0059] The task allocation module includes a task analysis unit and a task matching unit; the task analysis unit is used to obtain the corresponding regional task type, the number of task types and the processing technology sub-parameters corresponding to the task type through the regional task classification model according to the corresponding regional processing technology sub-parameters after regional division; the task matching unit is used to obtain the regional task execution list corresponding to the nozzle through the matching algorithm according to the obtained corresponding regional task type, the processing technology sub-parameters corresponding to the task type and the nozzle properties.

[0060] Specifically, the printing control module includes a path generation unit and a collaborative control unit;

[0061] A path generation unit is used to generate a printing path of each nozzle corresponding to the task type in each area through a path generation algorithm according to the processing technology sub-parameters corresponding to each area and the task type corresponding to each nozzle in each area, so as to obtain a multi-nozzle segmented area path;

[0062] A collaborative control unit is used to build a multi-agent collaborative control model and integrate the multi-agent collaborative control model with the multi-nozzle segmented area path to control the dynamic 3D printing process of the multi-nozzle in each area in real time;

[0063] The monitoring feedback module includes an error unit, an error prediction unit and a pre-adjustment unit;

[0064] The error prediction unit is used to obtain the printing index deviation according to the error unit, and predict the printing index prediction deviation at the next moment through the built-in error trend prediction model;

[0065] The pre-adjustment unit is used to feed back the prediction deviation of the printing index at the next moment to the multi-agent collaborative control model, and make real-time pre-adjustments to the errors of each nozzle.

[0066] A computer-readable storage medium stores computer instructions, which, when executed, execute a multi-nozzle collaborative control method for 3D printing.

[0067] Compared with the prior art, the present invention has the following beneficial effects:

[0068] In view of the deficiencies in the prior art, the present invention constructs a three-dimensional parameter model of the workpiece to be printed through three-dimensional software and performs detailed stratification and area division, ensuring that each nozzle can obtain accurate processing technology sub-parameters, and then efficiently assigns tasks to the nozzles through a classification matching algorithm. Secondly, an advanced path algorithm is used to generate a multi-nozzle segmented area path, which not only reduces the mutual interference between the nozzles, but also improves the continuity and smoothness of the printing process. Thirdly, a multi-agent collaborative control model is constructed through a reinforcement learning method. The model can dynamically adjust the nozzle path according to the real-time status to achieve efficient collaborative operation between multiple nozzles. In addition, the system also integrates non-destructive testing sensors, which can monitor printing indicators in real time during the printing process, and predict deviations in advance through an error trend prediction model, and adjust nozzle parameters in time, effectively reducing printing errors. BRIEF DESCRIPTION OF THE DRAWINGS

[0069] Figure 1 This is a flow chart of a multi-nozzle collaborative control method for 3D printing according to Embodiment 1 of the present invention;

[0070] Figure 2 This is a diagram of the regional division process architecture of Example 1 of the present invention;

[0071] Figure 3 This is a module diagram of the multi-nozzle collaborative control system for 3D printing of the present invention. DETAILED DESCRIPTION

[0072] Example 1

[0073] See also Figure 1 An embodiment of the present invention provides a multi-nozzle collaborative control method for 3D printing, which is applied to the 3D printing process of jewelry, comprising the following steps:

[0074] S1. Obtain processing parameters of the workpiece to be printed, construct a three-dimensional parameter model of the workpiece to be printed by using three-dimensional software, and pre-layer and divide the constructed three-dimensional parameter model into regions, and obtain processing sub-parameters of the corresponding regions after the region division;

[0075] Further, in this embodiment, the processing parameters of the part to be printed include material properties of the printing material, such as metal powder type, resin type, etc., layer thickness, printing speed, filling density, support structure and surface treatment parameters;

[0076] For further information, see Figure 2 , the steps of obtaining the corresponding regional processing technology sub-parameters after regional division include:

[0077] S101, setting the number of pre-layers M and the layer thickness, dividing the three-dimensional parameter model into layers according to the preset number of layers M and the layer thickness through the software, obtaining a single printing layer, analyzing the geometric parameters and the required printing material parameters of the divided single printing layer through the 3D software, and obtaining a regional attribute parameter set and a processing technology parameter set for each layer after the division;

[0078] S102, inputting each acquired single-layer region attribute parameter set and processing technology parameter set into the grid algorithm to obtain each single-layer divided region sequence and the corresponding divided region processing technology sub-parameter set;

[0079] S103, setting a printing order of a single-layer printing layer, and setting a starting printing area in each obtained single-layer divided area sequence, and sorting the single-layer divided area sequence in ascending order according to the distance value between the center point of each divided area and the center point of the starting printing area position, and labeling the area, obtaining a printing area sequence sequence, and marking the processing technology sub-parameter set of the corresponding area under each divided area in the obtained printing area sequence sequence; in this embodiment, the smaller the distance value between the center point of each divided area and the center point of the starting printing area position, the earlier the printing order is, and the earlier it is printed;

[0080] S104, repeat S102 and S103 to obtain the M-layer division area space of the three-dimensional parameter model [ d mn ] M × N and divide the regional processing technology sub-parameter space, where represents the nth partition area in the mth layer, and N represents the total number of partition areas in a single layer;

[0081] S2. According to the obtained processing sub-parameters of the corresponding area and the nozzle properties, the corresponding divided area printing task types and the regional task execution list corresponding to the nozzle are obtained through a classification matching algorithm; further, the specific steps of S2 in this embodiment include:

[0082] S201, obtain S104 [ d mn ] M × N And the corresponding partitioned regional processing technology sub-parameter space is input into the pre-trained regional task classification model to obtain The 3D printing task type space corresponding to the divided area [ d nm p ] M × N × P , the number of task types and the processing technology sub-parameter space corresponding to the task type [ d nm pk ] M × N × K ,in, represents the pth 3D printing task type in the nth partition area in the mth layer, represents the kth processing sub-parameter corresponding to the pth 3D printing task type in the nth divided area in the mth layer, K represents the total number of processing sub-parameters corresponding to the pth 3D printing task type; P represents the total number of task types in the nth divided area in the mth layer; further, the regional task classification model in this embodiment is constructed by a random forest algorithm, and is obtained through historical [ d mn ] M × N and the corresponding divided area processing technology sub-parameter space is trained to obtain; further, in this embodiment, the jewelry printing task types include fine printing, fast filling, support structure printing, texture printing, detail enhancement printing, surface smoothing printing, special material printing and color gradient printing;

[0083] S202: Obtain nozzle attribute parameters, and use the obtained nozzle attribute parameters to construct a nozzle parameter space [ u j q ] J × Q ,in, represents the qth attribute parameter of the jth nozzle, J represents the total number of nozzles, and Q represents the total number of attribute parameters corresponding to each nozzle; the nozzle attribute parameters include nozzle diameter, nozzle temperature range, maximum printing speed, minimum layer thickness, maximum layer thickness, heating rate, cooling rate, nozzle positioning accuracy, nozzle spray volume and nozzle shape;

[0084] S203, the acquired 3D printing task type space [ d nm p ] M × N × P Corresponding processing technology sub-parameter space [ d nm pk ] M × N × K and nozzle parameter space [ u j q ] J × Q , input into the matching algorithm to obtain the jth nozzle and The matching degree of task types ,in Indicates the matching degree between the pth task type and the jth sprinkler in the nth partition area in the mth layer;

[0085] S204: Setting matching threshold ,when Greater than Then the pth 3D printing task type in the nth partition area in the mth layer is assigned to the jth nozzle, otherwise it is not assigned to the jth nozzle;

[0086] S205, repeat S204 process, from [ d nm p ] M × N × P Extracting satisfaction from Task type, construct the printing task set of the j-th nozzle [ b nm v ] V , It indicates that the j-th print head corresponds to the v-th print task in the print task set, and the v-th print task is located in the n-th partition area in the m-th layer. V indicates the total number of tasks in the print task set corresponding to the j-th print head.

[0087] S206, sorting the tasks in the print task set of the j-th nozzle according to the print area sequence obtained in S103, and making a one-to-one correspondence between the sorted tasks and the area numbers in the print area sequence, to obtain a print task sequence set of the j-th nozzle;

[0088] S207, repeat the process of S206 to obtain the printing task sequence space corresponding to J nozzles;

[0089] S208, creating a print task list, and embedding the obtained print task sequence space corresponding to the J nozzles and the processing technology sub-parameter corresponding to each task in the print task sequence space into the print task list, and obtaining a regional task execution list corresponding to the J nozzles;

[0090] The process first classifies the task type of each divided area by using the pre-trained regional task classification model to ensure that each task type can be matched with the most suitable nozzle. Subsequently, through the matching algorithm and setting the matching degree threshold, efficient matching between the nozzle and the task type is achieved, thereby ensuring that each nozzle can undertake the printing task that best suits its own characteristics. In addition, by sorting the printing tasks and assigning tasks according to the order sequence of the printing area, the movement distance of the nozzle between different tasks is significantly reduced, and the printing efficiency is improved. Finally, by creating a printing task list and embedding the task sequence and processing technology sub-parameters into it, the comprehensive optimization of the multi-nozzle 3D printing process is achieved, which ensures the printing quality and efficiency while also improving the flexibility and adaptability of the system.

[0091] S3, initialize the multi-nozzle position, and generate the multi-nozzle segmented area path through the path generation algorithm according to the processing technology sub-parameters corresponding to each area, the printing task type and the area task execution list corresponding to the nozzle;

[0092] Furthermore, the specific steps of initializing the multi-nozzle position in this embodiment include:

[0093] a. Set the center point of the printing platform as the coordinate origin of the three-dimensional coordinate system, and project the obtained three-dimensional parameter model M-layer divided area space into the constructed three-dimensional coordinate system;

[0094] b. Set the first starting printing area of ​​the M-layer division area space of the three-dimensional parameter model, obtain the angle deviation of the J nozzles through the arc cosine trigonometric function according to the current position of the J nozzles and the first starting printing area position, and use the calculated angle deviation to correct and initialize the current position of the J nozzles;

[0095] Furthermore, in this embodiment, the step of acquiring the multi-nozzle segmented area path includes:

[0096] S301, inputting the printing task sequence set and corresponding processing technology sub-parameters of the j-th nozzle in the regional task execution list obtained in S208 into the path generation algorithm built into the 3D software to generate a task execution path corresponding to the j-th nozzle;

[0097] S302: insert V task start execution time points into the task execution path corresponding to the jth nozzle according to the print task sequence set corresponding to the jth nozzle and the corresponding processing technology sub-parameters. , the specific start time of task execution of the vth task is:

[0098]

[0099] in, Indicates the start time of the first task execution of the j-th nozzle, Indicates the starting execution time point of the v-1th task corresponding to the jth nozzle, Indicates the execution time length of the v-1th task of the jth sprinkler; Indicates the starting waiting time for the jth nozzle to perform the vth task;

[0100] S303, collect the historical start waiting time sequence and the corresponding printing index deviation of the j-th nozzle when executing the task in the printing task sequence set, use the historical start waiting time sequence and the corresponding printing index deviation to train the waiting time deviation prediction model constructed by the support vector machine, and obtain the starting waiting time deviation of the v-th task of the j-th nozzle ;

[0101] S304, feeding back the calculated starting waiting time deviation of the vth task printing task of the jth nozzle to the task starting execution time point of the vth task, adjusting the starting execution time point of the vth task execution corresponding to the jth nozzle, and obtaining the adjusted task starting execution time point of the vth task, specifically:

[0102]

[0103] S305, repeating the processes of S303 and S304 to adjust the task start execution time points of the V tasks corresponding to the j-th nozzle, and using the adjusted task start execution time points to segment the generated task execution path, to obtain the segmented area path corresponding to the j-th nozzle;

[0104] S306. Repeat the process of S301-S305 to obtain the segmented regional paths corresponding to the J nozzles, and save the segmented regional paths corresponding to the J nozzles into the regional task execution list to obtain the regional task path execution list.

[0105] This process not only generates the optimal task execution path for each nozzle by inputting the printing task sequence set and processing technology sub-parameters into the path generation algorithm, but also effectively reduces unnecessary waiting time by dynamically adjusting the task start execution time point, thereby improving the overall printing efficiency. In addition, the support vector machine model trained with historical data predicts the waiting time deviation and adjusts the task start time accordingly, further reducing the printing index deviation caused by waiting and ensuring the consistency and stability of printing.

[0106] S4. Build a multi-agent collaborative control model through reinforcement learning, and build the multi-nozzle collaborative control strategy into the multi-nozzle segmented area path to control the dynamic 3D printing state of the multi-nozzle corresponding path in real time;

[0107] Furthermore, in this embodiment, the multi-agent collaborative control model includes J single control sub-models and one collaborative control sub-model, and the J single control sub-models correspond to the J sprinklers one by one; the specific steps of constructing the single control sub-model include:

[0108] S401, construct the current input state according to the segmented regional path of the j-th nozzle, the printing task sequence set and the processing sub-parameters corresponding to the printing task saved in the regional task path execution list of S306:

[0109] ,

[0110] in, Indicates the position of the j-th nozzle in the divided area at the current time t, It indicates the progress of the task execution in the printing task sequence corresponding to the j-th nozzle at the current time t, that is, the task completion amount. It indicates the moving speed of the jth nozzle when performing the vth task at the current time t. It indicates the printing speed of the jth nozzle when executing the vth task at the current time t. It indicates the thickness corresponding to the task v performed by the j-th nozzle at the current time t. Indicates the starting time point corresponding to the jth nozzle executing the vth task, Indicates the printing index deviation corresponding to the j-th print head executing the v-th task at the current time t;

[0111] S402, setting the current moment execution trigger information based on the set current moment input state, specifically:

[0112] when The first execution action information is triggered when the corresponding execution adjustment and Actions, represents the processing thickness threshold corresponding to the vth task performed by the jth nozzle;

[0113] when The second execution action information is triggered when a single or combined execution adjustment is performed. , and Actions;

[0114] When the order of executing tasks corresponding to the j-th print head does not satisfy the order in the printing task sequence set, the third execution action information is triggered, that is, the action of adjusting the task execution sequence is executed;

[0115] S403: Construct the current execution action according to the current execution trigger information ,in, Indicates the action that triggers the first execution action information to be executed. Indicates the action that triggers the second execution action information to be executed. Indicates the action executed by triggering the third execution action information;

[0116] S404. Set the reward function according to the action executed at the current moment of the construction:

[0117]

[0118] in, It indicates the thickness error corresponding to the j-th nozzle performing the v-th task at the current time t, represents the reward error corresponding to the o-th adjustment of the task execution order, O represents the total number of times the j-th nozzle completes the task adjustment execution order in the corresponding printing task sequence set, is the gain coefficient corresponding to the reward function.

[0119] S405: construct a single control sub-model based on the SAC algorithm, and input the current input state, current execution trigger information, current execution action and reward function into the single control sub-model for training, and output the adjustment amount of the execution action corresponding to the j-th nozzle trigger information, and perform real-time control adjustment on the task execution process of the j-th nozzle;

[0120] S406, real-time monitoring to obtain the real-time processing parameters of J nozzles under the control of J single control sub-models and the movement parameters of the nozzles when performing tasks, and input the obtained movement parameters of the nozzles when performing tasks into the collision prediction model constructed by the support vector machine to obtain the probability sequence of the collision between the j-th nozzle and the g-th nozzle among the J nozzles at the current moment;

[0121] S407, setting a collision probability threshold, comparing the obtained collision probability sequence at the current moment with the collision probability threshold, and obtaining the nozzles and the corresponding collision probabilities corresponding to the collision probability at the current moment being greater than the collision probability threshold;

[0122] S408, feeding back the nozzles corresponding to the current moment collision probability greater than the collision probability threshold and the corresponding collision probability to the cooperative control sub-model constructed by the control algorithm, generating control signals for the nozzles corresponding to the current moment collision probability greater than the collision probability threshold, and transmitting the generated control signals to the single control sub-model of the corresponding nozzles to control the nozzles to perform collision avoidance actions;

[0123] S409, real-time monitoring of the real-time layer thickness parameters of the m-th single-layer printing layer completed by the nozzle, and comparing the real-time layer thickness parameters of the m-th single-layer printing layer with the layer thickness parameters in the processing parameter set of each layer after division obtained in S101. If they are not equal, the layer thickness error calculated by the real-time layer thickness parameters and the layer thickness parameters in the processing parameter set is fed back to the collaborative control sub-model to generate a collaborative control adjustment signal;

[0124] S410, feeding back the generated collaborative control adjustment signal to the single control sub-model corresponding to all nozzles executing the m-th single-layer printing task, triggering the model to perform the adjustment , or The action makes the real-time layer thickness parameter equal to the layer thickness parameter in the processing parameter set;

[0125] This process builds a single control sub-model for each nozzle, combines the current task status and environmental changes, and automatically adjusts the nozzle speed, printing speed, thickness and other key parameters to ensure the smooth completion of the printing task; secondly, by setting the collision prediction model and collision probability threshold, the system can timely identify and handle possible collision events, reducing the risk of equipment damage; thirdly, by real-time monitoring of the thickness parameters of each printing layer and comparing it with the predetermined target value, the dimensional accuracy and structural integrity of the printed part are ensured, thereby improving the quality of the final product; the dynamic adjustment mechanism of this process greatly optimizes the performance of multi-nozzle collaborative work, which not only improves printing efficiency, but also ensures the safety of the printing process and the accuracy of the printing results;

[0126] S5. Use non-destructive testing sensors to monitor the 3D printing process of each nozzle in real time, obtain the printing index data of the workpiece to be printed, obtain the printing index deviation according to the obtained printing index data and the processing parameters of the workpiece to be printed, and input the obtained printing index deviation into the built-in error trend prediction model to obtain the printing index prediction deviation of each nozzle at the next moment; further, the specific steps of constructing the error trend prediction model in this embodiment are:

[0127] S501, obtaining historical processing parameters of J nozzles and the printing index deviation sequence corresponding to each nozzle, and performing preprocessing; further, the printing index data in this embodiment is the real-time obtained jewelry geometric processing parameters, such as the accuracy of detail features such as height, thickness, text on jewelry, cavities and bending angles.

[0128] S502, constructing J error trend prediction sub-models through support vector machines, and integrating the J error trend prediction sub-models into the error trend prediction model through an integration algorithm, and simultaneously inputting the pre-processed J nozzle historical processing parameters and the printing index deviation sequence corresponding to each nozzle into each error trend prediction sub-model, training the error trend prediction model, and obtaining the trained error trend prediction model;

[0129] S503, embed the trained error trend prediction model into the segmented area paths corresponding to the J nozzles, and input the actual processing parameters at the current moment obtained by real-time monitoring into the error trend prediction model to predict the printing index deviation at the next moment;

[0130] S6. Feedback the predicted deviation of the printing index of each nozzle at the next moment into the multi-agent collaborative control model, and make pre-adjustments to each nozzle during the printing process so that the printing index data of the to-be-printed part is equal to the processing parameters of the to-be-printed part in real time.

[0131] Example 2

[0132] See also Figure 3 , another embodiment provided by the present invention: a multi-nozzle collaborative control system for 3D printing, comprising: a three-dimensional module, a task allocation module, a printing control module and a monitoring feedback module;

[0133] The three-dimensional module is used to construct a three-dimensional model of the part to be printed and perform regional division; the three-dimensional module includes a three-dimensional construction unit and a regional division unit; the three-dimensional construction unit is used to construct a three-dimensional parameter model of the part to be printed through the three-dimensional construction software according to the acquired processing parameters of the part to be printed; the regional division unit is used to pre-layer the three-dimensional parameter model, and use the grid algorithm to perform virtual regional division and annotation on each layer after pre-layering, and obtain the corresponding regional processing sub-parameters after regional division;

[0134] The task allocation module is used to allocate specific printing tasks to each nozzle; the task allocation module includes a task analysis unit and a task matching unit;

[0135] The task analysis unit is used to obtain the task type, the number of task types and the processing sub-parameters corresponding to the task type according to the processing sub-parameters of the corresponding area after the area division through the regional task classification model; the task matching unit is used to obtain the regional task execution list corresponding to the nozzle through the matching algorithm according to the obtained task type, the processing sub-parameters corresponding to the task type and the nozzle properties;

[0136] A printing control module is used to generate a multi-nozzle segmented area path and use a built-in multi-intelligent collaborative control model to control the multi-nozzle printing process; the printing control module includes an initialization unit, a path generation unit and a collaborative control unit;

[0137] An initialization unit, used to set a printing coordinate system and a printing starting point, and to initialize and calibrate the position of each printing nozzle according to the set printing coordinate system, printing starting point, three-dimensional parameter model and current position of the printing nozzle;

[0138] A path generation unit is used to generate a printing path of each nozzle corresponding to the task type in each area through a path generation algorithm according to the processing technology sub-parameters corresponding to each area and the task type corresponding to each nozzle in each area, so as to obtain a multi-nozzle segmented area path;

[0139] A collaborative control unit is used to build a multi-agent collaborative control model and integrate the multi-agent collaborative control model with the multi-nozzle segmented area path to control the dynamic 3D printing process of the multi-nozzle in each area in real time;

[0140] A monitoring and feedback module is used to monitor the printing index deviation of the printing process in real time, and use the obtained printing index deviation to make pre-adjustments to the printing process; the monitoring and feedback module includes an error unit, an error prediction unit and a pre-adjustment unit;

[0141] An error unit is used to monitor and obtain the printing index data of the workpiece to be printed in real time, and obtain the printing index deviation according to the obtained printing index data and the processing parameters of the workpiece to be printed;

[0142] The error prediction unit is used to predict the printing index prediction deviation at the next moment through the built-in error trend prediction model according to the obtained printing index deviation;

[0143] The pre-adjustment unit is used to feed back the prediction deviation of the printing index at the next moment to the multi-agent collaborative control model, and make real-time pre-adjustments to the errors of each nozzle.

[0144] Example 3

[0145] A computer-readable storage medium stores computer instructions, which, when executed, execute a multi-nozzle collaborative control method for 3D printing.

[0146] An electronic device includes a memory and a processor, wherein the memory stores a computer program, and the processor implements a multi-nozzle collaborative control method for 3D printing when executing the computer program.

[0147] The embodiments of the present invention are described above in conjunction with the accompanying drawings, but the present invention is not limited to the above-mentioned specific implementation modes, which are merely illustrative rather than restrictive. Under the guidance of the present invention, ordinary technicians in the field may also change, modify, replace and modify the above-mentioned embodiments without departing from the scope of protection of the purpose of the present invention and the claims, and all of these are within the protection of the present invention.

[0148] If the disclosed technical solution involves personal information, the product using the disclosed technical solution has clearly informed the personal information processing rules and obtained the individual's voluntary consent before processing the personal information. If the disclosed technical solution involves sensitive personal information, the product using the disclosed technical solution has obtained the individual's separate consent before processing the sensitive personal information, and at the same time meets the "explicit consent" requirement. For example, on personal information collection devices such as cameras, clear and prominent signs are set to inform that the personal information collection scope has been entered and personal information will be collected. If the individual voluntarily enters the collection scope, it is deemed that he or she agrees to the collection of his or her personal information; or on the device that processes personal information, when the personal information processing rules are notified by obvious signs / information, the individual's authorization is obtained through pop-up information or by asking the individual to upload his or her personal information; among them, the personal information processing rules may include information such as the personal information processor, the purpose of personal information processing, the processing method, and the type of personal information processed.

Claims

1. A multi-nozzle collaborative control method for 3D printing, characterized in that: include: S1. Obtain processing parameters of the workpiece to be printed, construct a three-dimensional parameter model of the workpiece to be printed by using three-dimensional software, and pre-layer and divide the constructed three-dimensional parameter model into regions, and obtain processing sub-parameters of the corresponding regions after the region division; S2. According to the obtained processing sub-parameters of the corresponding area and the nozzle properties, the printing task types of the corresponding divided areas and the regional task execution list corresponding to the nozzle are obtained through a classification matching algorithm; S3, initialize the multi-nozzle position, and generate the multi-nozzle segmented area path through the path generation algorithm according to the processing technology sub-parameters corresponding to each area, the printing task type and the area task execution list corresponding to the nozzle; S4. Construct a multi-agent collaborative control model through reinforcement learning, and embed the multi-nozzle collaborative control strategy into the multi-nozzle segmented area path to control the dynamic 3D printing state of the multi-nozzle corresponding path in real time; S5. Use non-destructive testing sensors to monitor the 3D printing process of each nozzle in real time, obtain the printing index data of the workpiece to be printed, obtain the printing index deviation according to the obtained printing index data and the processing parameters of the workpiece to be printed, and input the obtained printing index deviation into the built-in error trend prediction model to obtain the printing index prediction deviation of each nozzle at the next moment; S6, feeding back the predicted deviation of the printing index of each nozzle at the next moment to the multi-agent collaborative control model, and making pre-adjustments to each nozzle during the printing process, so that the printing index data of the to-be-printed part is equal to the processing parameters of the to-be-printed part in real time; The step of obtaining the processing sub-parameters of the corresponding regions after the region division in S1 includes: S101, setting the number of pre-layers M and the layer thickness, dividing the three-dimensional parameter model into layers according to the preset number of layers M and the layer thickness through the software, obtaining a single printing layer, analyzing the geometric parameters and the required printing material parameters of the divided single printing layer through the 3D software, and obtaining a regional attribute parameter set and a processing technology parameter set for each layer after the division; S102, inputting each acquired single-layer region attribute parameter set and processing technology parameter set into the grid algorithm to obtain each single-layer divided region sequence and the corresponding divided region processing technology sub-parameter set; S103, setting a printing order of a single-layer printing layer, and setting a starting printing area in each obtained single-layer divided area sequence, and sorting the single-layer divided area sequence in ascending order according to the distance value between the center point of each divided area and the center point of the starting printing area position, obtaining a printing area sequence sequence, and marking the processing technology sub-parameter set of the corresponding area under each divided area in the obtained printing area sequence sequence; S104, repeat S102 and S103 to obtain the M-layer division area space of the three-dimensional parameter model and divide the regional processing technology sub-parameter space, where represents the nth partition area in the mth layer, and N represents the total number of partition areas in a single layer; The specific steps of S2 include: S201, obtain S104 And the corresponding partitioned regional processing technology sub-parameter space is input into the pre-trained regional task classification model to obtain The 3D printing task type space corresponding to the divided area , the number of task types and the processing technology sub-parameter space corresponding to the task type ,in, represents the pth 3D printing task type in the nth partition area in the mth layer, represents the kth processing sub-parameter corresponding to the pth 3D printing task type in the nth divided area in the mth layer, K represents the total number of processing sub-parameters corresponding to the pth 3D printing task type; P represents the total number of task types in the nth divided area in the mth layer; S202: Obtain nozzle attribute parameters, and use the obtained nozzle attribute parameters to construct a nozzle parameter space ,in, represents the qth attribute parameter of the jth nozzle, J represents the total number of nozzles, and Q represents the total number of attribute parameters corresponding to each nozzle; S203, the acquired 3D printing task type space Corresponding processing technology sub-parameter space and nozzle parameter space , input into the matching algorithm to obtain the jth nozzle and The matching degree of task types ,in Indicates the matching degree between the pth task type and the jth sprinkler in the nth partition area in the mth layer; S204: Setting matching threshold ,when Greater than Then the pth 3D printing task type in the nth partition area in the mth layer is assigned to the jth nozzle, otherwise it is not assigned to the jth nozzle; S205, repeat S204 process, from Extracting satisfaction from Task type, construct the printing task set of the j-th nozzle , It indicates that the j-th print head corresponds to the v-th print task in the print task set, and the v-th print task is located in the n-th partition area in the m-th layer, and V indicates the total number of tasks in the print task set corresponding to the j-th print head; The specific steps of S2 also include: S206, sorting the tasks in the print task set of the j-th nozzle according to the print area sequence obtained in S103, and making a one-to-one correspondence between the sorted tasks and the area numbers in the print area sequence, to obtain a print task sequence set of the j-th nozzle; S207, repeat the process of S206 to obtain the printing task sequence space corresponding to J nozzles; S208, creating a print task list, and embedding the obtained print task sequence space corresponding to the J nozzles and the processing technology sub-parameters corresponding to each task in the print task sequence space into the print task list, and obtaining the regional task execution list corresponding to the J nozzles.

2. The multi-nozzle collaborative control method for 3D printing according to claim 1, characterized in that: The step of acquiring the multi-nozzle segmented area path comprises: S301, inputting the printing task sequence set and corresponding processing technology sub-parameters of the j-th nozzle in the regional task execution list obtained in S208 into the path generation algorithm built into the 3D software to generate a task execution path corresponding to the j-th nozzle; S302: insert V task start execution time points into the task execution path corresponding to the jth nozzle according to the print task sequence set and the corresponding processing technology sub-parameters corresponding to the jth nozzle. , the specific start time of task execution of the vth task is: ； in, Indicates the start time of the first task execution of the j-th nozzle, Indicates the starting execution time point of the v-1th task corresponding to the jth nozzle, Indicates the execution time length of the v-1th task of the jth sprinkler; Indicates the starting waiting time for the jth nozzle to perform the vth task; S303, collect the historical start waiting time sequence and the corresponding printing index deviation of the j-th nozzle when executing the task in the printing task sequence set, use the historical start waiting time sequence and the corresponding printing index deviation to train the waiting time deviation prediction model constructed by the support vector machine, and obtain the starting waiting time deviation of the v-th task of the j-th nozzle ; S304, feeding back the calculated starting waiting time deviation of the vth task printing task of the jth nozzle to the task starting execution time point of the vth task, adjusting the starting execution time point of the vth task execution corresponding to the jth nozzle, and obtaining the adjusted task starting execution time point of the vth task, specifically: ； S305, repeating the processes of S303 and S304 to adjust the task start execution time points of the V tasks corresponding to the j-th nozzle, and using the adjusted task start execution time points to segment the generated task execution path, to obtain the segmented area path corresponding to the j-th nozzle; S306. Repeat the process of S301-S305 to obtain the segmented regional paths corresponding to the J nozzles, and save the segmented regional paths corresponding to the J nozzles into the regional task execution list to obtain the regional task path execution list.

3. The multi-nozzle collaborative control method for 3D printing according to claim 2, characterized in that: The multi-agent collaborative control model includes J single control sub-models and one collaborative control sub-model, and the J single control sub-models correspond to the J sprinklers one by one; The specific steps of constructing the single control sub-model include: S401, construct the current input state according to the segmented regional path of the j-th nozzle, the printing task sequence set and the processing sub-parameters corresponding to the printing task saved in the regional task path execution list of S306: ， in, Indicates the position of the j-th nozzle in the divided area at the current time t, Indicates the progress of the task execution in the printing task sequence set corresponding to the j-th nozzle at the current time t. It indicates the moving speed of the jth nozzle when performing the vth task at the current time t. It indicates the printing speed of the jth nozzle when executing the vth task at the current time t. It indicates the thickness corresponding to the task v performed by the j-th nozzle at the current time t. Indicates the starting time point corresponding to the jth nozzle executing the vth task, Indicates the printing index deviation corresponding to the j-th print head executing the v-th task at the current time t; S402, setting the current moment execution trigger information based on the set current moment input state, specifically: when The first execution action information is triggered when the corresponding execution adjustment and Actions, represents the processing thickness threshold corresponding to the vth task performed by the jth nozzle; when The second execution action information is triggered when a single or combined execution adjustment is performed. , and Actions; When the order of executing tasks corresponding to the j-th print head does not satisfy the order in the printing task sequence set, the third execution action information is triggered, that is, the action of adjusting the task execution sequence is executed; S403: Construct the current execution action according to the current execution trigger information ,in, Indicates the action that triggers the first execution action information to be executed. Indicates the action that triggers the second execution action information to be executed. Indicates the action executed by triggering the third execution action information; S404. Set the reward function according to the action executed at the current moment of the construction: ； in, It indicates the thickness error corresponding to the j-th nozzle performing the v-th task at the current time t. represents the reward error corresponding to the o-th adjustment of the task execution order, O represents the total number of times the j-th nozzle completes the task adjustment execution order in the corresponding printing task sequence set, is the gain coefficient corresponding to the reward function.

4. The multi-nozzle collaborative control method for 3D printing according to claim 3, characterized in that: The specific steps of the collaborative control sub-model include: S405, real-time monitoring to obtain the real-time processing parameters of J nozzles under the control of J single control sub-models and the movement parameters of the nozzles when performing tasks, and input the obtained movement parameters of the nozzles when performing tasks into the collision prediction model constructed by the support vector machine to obtain the probability sequence of the collision between the j-th nozzle and the g-th nozzle among the J nozzles at the current moment; S406, setting a collision probability threshold, comparing the obtained collision probability sequence at the current moment with the collision probability threshold, and obtaining the nozzles and the corresponding collision probabilities corresponding to the collision probability at the current moment being greater than the collision probability threshold; S407, feeding back the nozzles corresponding to the current moment collision probability greater than the collision probability threshold and the corresponding collision probability to the cooperative control sub-model constructed by the control algorithm, generating control signals for the nozzles corresponding to the current moment collision probability greater than the collision probability threshold, and transmitting the generated control signals to the single control sub-model of the corresponding nozzles to control the nozzles to perform collision avoidance actions; S408, real-time monitoring of the real-time layer thickness parameters of the m-th single-layer printing layer completed by the nozzle, and comparing the real-time layer thickness parameters of the m-th single-layer printing layer with the layer thickness parameters in the processing parameter set of each layer after division obtained in S101. If they are not equal, the layer thickness error calculated by the real-time layer thickness parameters and the layer thickness parameters in the processing parameter set is fed back to the collaborative control sub-model to generate a collaborative control adjustment signal; S409: Feedback the generated collaborative control adjustment signal to the single control sub-model corresponding to all nozzles executing the m-th single-layer printing task, triggering the model to perform the adjustment , or The action makes the real-time layer thickness parameter equal to the layer thickness parameter in the machining process parameter set.

5. A multi-nozzle collaborative control system for 3D printing, which is implemented based on the multi-nozzle collaborative control method for 3D printing according to any one of claims 1 to 4, characterized in that: include: Three-dimensional module, task allocation module, printing control module and monitoring feedback module; The three-dimensional module includes a three-dimensional construction unit and a region division unit; the region division unit is used to pre-layer the three-dimensional parameter model constructed by the three-dimensional construction unit, and use a grid algorithm to perform virtual region division and annotation on each layer after the pre-layering, and obtain the corresponding regional processing technology sub-parameters after the regional division; The task assignment module includes a task analysis unit and a task matching unit; the task analysis unit is used to obtain the corresponding regional task type, the number of task types and the processing technology sub-parameters corresponding to the task type through the regional task classification model according to the corresponding regional processing technology sub-parameters after regional division; the task matching unit is used to obtain the regional task execution list corresponding to the nozzle through the matching algorithm according to the acquired corresponding regional task type, the processing technology sub-parameters corresponding to the task type and the nozzle properties.

6. The multi-nozzle collaborative control system for 3D printing according to claim 5, characterized in that: The printing control module includes a path generation unit and a collaborative control unit; The path generation unit is used to generate a printing path for each nozzle corresponding to the task type in each area through a path generation algorithm according to the processing technology sub-parameters corresponding to each area and the task type corresponding to each nozzle in each area, so as to obtain a multi-nozzle segmented area path; The collaborative control unit is used to construct a multi-agent collaborative control model, and integrate the multi-agent collaborative control model with the multi-nozzle segmented area path to control the dynamic 3D printing process of the multi-nozzle in each area in real time; The monitoring feedback module includes an error unit, an error prediction unit and a pre-adjustment unit; The error prediction unit is used to obtain the printing index deviation according to the error unit, and predict the printing index prediction deviation at the next moment through the built-in error trend prediction model; The pre-adjustment unit is used to feed back the prediction deviation of the printing index at the next moment to the multi-agent collaborative control model, and perform real-time pre-adjustment of the error of each nozzle.

7. A computer-readable storage medium, characterized in that: Computer instructions are stored thereon, and when the computer instructions are executed, the multi-nozzle collaborative control method for 3D printing described in any one of claims 1 to 4 is executed.

Images