阅读说明：本技术 一种电弧增材制造过程的自适应控制系统及方法 (Self-adaptive control system and method for electric arc additive manufacturing process ) 是由 蔡笑宇 董博伦 林三宝 李�权 王福德 赵衍华 范成磊 杨春利 于 2020-11-27 设计创作，主要内容包括：本发明公开了一种电弧增材制造过程的自适应控制系统及方法。数据采集模块通过前置式红外热成像仪采集熔池前方的温度信息并传输至控制模块,通过前置式相机采集熔池图像并传输至控制模块,通过前置式结构光传感器采集上一沉积层的轮廓信息并传输至控制模块；控制模块的数据处理器用于处理数据采集模块传输的数据信息,控制模块的前馈控制器基于对电弧能量和送丝速度进行前馈控制；控制模块的反馈控制器对电弧能量进行反馈控制；通信模块将控制模块的控制信号传输至电源和送丝机；控制器自学习模块用于前馈控制器人工智能模型的训练。本发明提高了电弧增材制造技术对复杂结构的适用性。(The invention discloses a self-adaptive control system and a self-adaptive control method in an electric arc additive manufacturing process. The data acquisition module acquires temperature information in front of a molten pool through a front-mounted infrared thermal imager and transmits the temperature information to the control module, the front-mounted camera acquires images of the molten pool and transmits the images to the control module, and the front-mounted structured light sensor acquires contour information of a previous settled layer and transmits the contour information to the control module; the data processor of the control module is used for processing data information transmitted by the data acquisition module, and the feedforward controller of the control module performs feedforward control on the arc energy and the wire feeding speed; a feedback controller of the control module performs feedback control on the arc energy; the communication module transmits the control signal of the control module to the power supply and the wire feeder; the controller self-learning module is used for training the feedforward controller artificial intelligence model. The invention improves the applicability of the electric arc additive manufacturing technology to complex structures.)

1. An adaptive control system for an electric arc additive manufacturing process is characterized by comprising a data acquisition module, a control module, a communication module and a controller learning module, wherein the acquisition module, the control module and the communication module are connected in sequence,

the data acquisition module is used for acquiring temperature information in front of a molten pool through a front-mounted infrared thermal imager and transmitting the temperature information to the control module, acquiring a molten pool image through a front-mounted camera and transmitting the molten pool image to the control module, and acquiring profile information of a previous deposition layer through a front-mounted structured light sensor and transmitting the profile information to the control module;

the control module comprises a data processor, a feedforward controller and a feedback controller, wherein the data processor is used for processing data information transmitted by the data acquisition module, extracting a temperature field and a temperature gradient field in front of a molten pool from temperature field information transmitted by the front infrared thermal imager, extracting the real-time size of the molten pool from a molten pool image transmitted by the front camera, and extracting the size of a previous deposition layer from profile information of the previous deposition layer transmitted by the front structured light sensor; the feedforward controller performs feedforward control on the arc energy according to the information of a temperature field and a temperature gradient field in front of a molten pool based on an artificial intelligence model, and performs feedforward control on the wire feeding speed according to the size of the previous deposition layer; the feedback controller performs feedback control on the arc energy according to the real-time size of the molten pool;

the communication module is used for transmitting the control signal of the control module to the power supply to realize the real-time regulation of the arc energy, and transmitting the control signal of the control module to the wire feeder to realize the real-time regulation of the wire feeding speed;

the controller self-learning module is used for training the artificial intelligence model of the feedforward controller.

2. The adaptive control system for an arc additive manufacturing process according to claim 1, wherein the feedforward controller is based on an artificial neural network that predicts an increase in arc energy required to reach the deposition process from information on a temperature field and a temperature gradient field in front of the molten pool, and predicts an increase in wire feed speed required to reach the deposition process from information on a size of a previous deposition layer.

3. The adaptive control system for the arc additive manufacturing process according to claim 1, wherein the controller self-learning module obtains a large amount of theoretical data of the arc additive manufacturing process temperature field with a random structure through finite element numerical simulation, and the artificial intelligence model of the feedforward controller achieves a desired accuracy through iterative learning.

4. The adaptive control method of an arc additive manufacturing process according to claim 1, based on the adaptive control system of an arc additive manufacturing process according to any one of claims 1 to 3, comprising the steps of:

before the electric arc additive manufacturing process starts, the controller self-learning module carries out iterative learning by using training data until the artificial intelligent model has certain precision, can accurately predict the electric arc energy increment required when the deposition process is carried out to the position according to the information of a temperature field and a temperature gradient field in front of a molten pool, and predicts the wire feeding speed increment required when the deposition process is carried out to the position according to the size information of the previous deposition layer;

in the electric arc additive manufacturing process, the data acquisition module acquires temperature information in front of a molten pool through a front-mounted infrared thermal imager and transmits the temperature information to the control module, a front-mounted camera acquires a molten pool image and transmits the molten pool image to the control module, and a front-mounted structured light sensor acquires profile information of a previous deposition layer and transmits the profile information to the control module;

step three, a data processor in the control module carries out real-time processing on data transmitted by the data acquisition module, extracts a temperature field and a temperature gradient field in front of a molten pool from temperature field information transmitted by a front-mounted infrared thermal imager, extracts the real-time size of the molten pool from a molten pool image transmitted by a front-mounted camera, and extracts the size of a previous deposition layer from profile information of the previous deposition layer transmitted by a front-mounted structured light sensor;

fourthly, a feedforward controller of the control module predicts the required arc energy increment when the deposition process is carried out to the position according to the information of the temperature field and the temperature gradient field in front of the molten pool, delays and outputs the arc energy increment to the communication module, predicts the required wire feeding speed increment when the deposition process is carried out to the position according to the size information of the previous deposition layer, and delays and outputs the arc energy increment to the communication module;

judging the arc energy increment required by ensuring the size precision according to the real-time size of the molten pool by a feedback controller of the control module, and outputting the arc energy increment to the communication module in real time;

and step six, the communication module transmits the control signal of the control module to the power supply to realize the real-time regulation of the arc energy, and transmits the control signal of the control module to the wire feeder to realize the real-time regulation of the wire feeding speed.

5. The adaptive control method for the arc additive manufacturing process according to claim 4, wherein in the first step, the controller self-learning module generates geometric models with large variable heat dissipation conditions by using a random algorithm, the geometric models are layered and subjected to path planning, then finite element numerical simulation is used for temperature field simulation, so that a temperature field in front of a molten pool, a temperature gradient field and an arc energy increment required for ensuring size stability when the molten pool moves to the place during arc additive manufacturing of each geometric model are obtained, the data are used for training an artificial intelligence model of the feedforward controller, and the artificial intelligence model of the feedforward controller achieves ideal precision through iterative learning.

Technical Field

The invention relates to a self-adaptive control system and a self-adaptive control method in an electric arc additive manufacturing process, and belongs to the technical field of electric arc additive manufacturing and intelligent manufacturing.

Background

The electric arc additive manufacturing is an additive manufacturing technology which takes an electric arc as a heat source and takes metal wire materials as filling materials. For a structure with larger size and complex geometric shape, the traditional equal-material manufacturing and material-reducing manufacturing process has the problems of complex manufacturing process, low flexibility degree, serious material waste and the like. By adopting the electric arc additive manufacturing technology, the manufacturing process can be simplified, the structural design of the product can be responded quickly, the manufacturing cost is reduced, and the application prospect is wide. The formation of the material in the additive manufacturing process is greatly affected by the change of heat dissipation conditions, and the dimensional accuracy is difficult to guarantee, which is one of the challenges facing the quality control of the additive manufacturing structure. At present, the manufacturing process of most additive manufacturing products is still open-loop, and the deposition parameters are determined in path planning. In industrial production, people hope to adjust process parameters in a self-adaptive manner according to the change of working conditions in the deposition process to ensure formation. However, the high complexity of the additive manufacturing process cannot be described by a universal mathematical model, and the difficulty in implementing process control is high, so that a feasible adaptive control method is proposed, which is a problem to be solved urgently at present. If adaptive control can be performed on the shaping of the material in the process of the electric arc additive manufacturing, the accuracy of the electric arc additive manufacturing is facilitated, and more stable product quality is obtained.

Disclosure of Invention

The invention provides a self-adaptive control system and a self-adaptive control method for an electric arc additive manufacturing process, aiming at the problems that the forming of materials is greatly influenced by the change of heat dissipation conditions and the dimensional accuracy is difficult to ensure in the electric arc additive manufacturing process. The problems of uneven forming and large deviation between the deposition size and the design size caused by variable heat dissipation conditions in the process of the electric arc additive manufacturing of the complex structure are fundamentally solved, and the applicability of the electric arc additive manufacturing technology to the complex structure is improved.

An adaptive control system for an arc additive manufacturing process comprises a data acquisition module, a control module, a communication module and a controller learning module, wherein the acquisition module, the control module and the communication module are sequentially connected,

the data acquisition module is used for acquiring temperature information in front of a molten pool through a front-mounted infrared thermal imager and transmitting the temperature information to the control module, acquiring a molten pool image through a front-mounted camera and transmitting the molten pool image to the control module, and acquiring profile information of a previous deposition layer through a front-mounted structured light sensor and transmitting the profile information to the control module;

the control module comprises a data processor, a feedforward controller and a feedback controller, wherein the data processor is used for processing data information transmitted by the data acquisition module, extracting a temperature field and a temperature gradient field in front of a molten pool from temperature field information transmitted by the front infrared thermal imager, extracting the real-time size of the molten pool from a molten pool image transmitted by the front camera, and extracting the size of a previous deposition layer from profile information of the previous deposition layer transmitted by the front structured light sensor; the feedforward controller performs feedforward control on the arc energy according to the information of a temperature field and a temperature gradient field in front of a molten pool based on an artificial intelligence model, and performs feedforward control on the wire feeding speed according to the size of the previous deposition layer; the feedback controller performs feedback control on the arc energy according to the real-time size of the molten pool;

the communication module is used for transmitting the control signal of the control module to the power supply to realize the real-time regulation of the arc energy, and transmitting the control signal of the control module to the wire feeder to realize the real-time regulation of the wire feeding speed;

the controller self-learning module is used for training the artificial intelligence model of the feedforward controller.

Further, the feedforward controller is based on an artificial neuron network, the artificial neuron network predicts the required arc energy increment when the deposition process is carried out to the position through the information of the temperature field and the temperature gradient field in front of the molten pool, and predicts the required wire feeding speed increment when the deposition process is carried out to the position through the information of the size of the last deposition layer.

Further, the controller self-learning module obtains a large amount of theoretical data of the temperature field in the arc additive manufacturing process of the random structure through finite element numerical simulation, and the artificial intelligent model of the feedforward controller achieves ideal precision through iterative learning.

An adaptive control method of an arc additive manufacturing process, based on the adaptive control system of the arc additive manufacturing process, the adaptive control method comprises the following steps:

before the electric arc additive manufacturing process starts, the controller self-learning module carries out iterative learning by using training data until the artificial intelligent model has certain precision, can accurately predict the electric arc energy increment required when the deposition process is carried out to the position according to the information of a temperature field and a temperature gradient field in front of a molten pool, and predicts the wire feeding speed increment required when the deposition process is carried out to the position according to the size information of the previous deposition layer;

in the electric arc additive manufacturing process, the data acquisition module acquires temperature information in front of a molten pool through a front-mounted infrared thermal imager and transmits the temperature information to the control module, a front-mounted camera acquires a molten pool image and transmits the molten pool image to the control module, and a front-mounted structured light sensor acquires profile information of a previous deposition layer and transmits the profile information to the control module;

step three, a data processor in the control module carries out real-time processing on data transmitted by the data acquisition module, extracts a temperature field and a temperature gradient field in front of a molten pool from temperature field information transmitted by a front-mounted infrared thermal imager, extracts the real-time size of the molten pool from a molten pool image transmitted by a front-mounted camera, and extracts the size of a previous deposition layer from profile information of the previous deposition layer transmitted by a front-mounted structured light sensor;

fourthly, a feedforward controller of the control module predicts the required arc energy increment when the deposition process is carried out to the position according to the information of the temperature field and the temperature gradient field in front of the molten pool, delays and outputs the arc energy increment to the communication module, predicts the required wire feeding speed increment when the deposition process is carried out to the position according to the size information of the previous deposition layer, and delays and outputs the arc energy increment to the communication module;

judging the arc energy increment required by ensuring the size precision according to the real-time size of the molten pool by a feedback controller of the control module, and outputting the arc energy increment to the communication module in real time;

and step six, the communication module transmits the control signal of the control module to the power supply to realize the real-time regulation of the arc energy, and transmits the control signal of the control module to the wire feeder to realize the real-time regulation of the wire feeding speed.

Further, in the first step, the controller self-learning module generates geometric models with large variable heat dissipation conditions by adopting a random algorithm, the geometric models are layered and subjected to path planning, finite element numerical simulation is adopted to perform temperature field simulation, an arc energy increment required for ensuring size stability when a temperature gradient field and a molten pool move to the position in front of a molten pool during arc additive manufacturing of each geometric model is obtained, the data are used for training of an artificial intelligence model of the feedforward controller, and the artificial intelligence model of the feedforward controller achieves ideal precision through iterative learning.

The main advantages of the invention are: the adaptive control system for the electric arc additive manufacturing process provided by the invention is not limited to specific materials, structural forms and electric arc additive manufacturing process methods, and has strong adaptability. By using infrared images, passive visual sensing and active visual sensing multi-information fusion input and feedforward and feedback composite control, the method has higher reliability for the electric arc additive manufacturing process with high nonlinearity and strong time-varying characteristics.

Drawings

FIG. 1 is a block diagram of an adaptive control system for an arc additive manufacturing process of the present invention;

FIG. 2 is a block diagram of a controller self-learning module of the present invention.

Detailed Description

The technical solutions in the embodiments of the present invention will be described clearly and completely with reference to the accompanying drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.

As shown in fig. 1, the present invention is an adaptive control system for an arc additive manufacturing process, comprising a data acquisition module, a control module, a communication module and a controller self-learning module; the data acquisition module acquires temperature information in front of a molten pool through a front-mounted infrared thermal imager and transmits the temperature information to the control module, acquires a molten pool image through a front-mounted camera and transmits the molten pool image to the control module, and acquires contour information of a previous deposition layer through a front-mounted structured light sensor and transmits the contour information to the control module; the control module comprises a data processor, a feedforward controller and a feedback controller; the data processor is used for processing the data information transmitted by the data acquisition module, extracting a temperature field and a temperature gradient field in front of a molten pool from the temperature field information transmitted by the infrared thermal imager, extracting the real-time size of the molten pool from a molten pool image transmitted by the camera through image processing, and extracting the size of a previous deposition layer from the profile information of the previous deposition layer transmitted by the structured light sensor through image processing; the feedforward controller predicts the required arc energy increment when the deposition process is carried out to the sampling positions of the temperature field and the temperature gradient field according to the information of the temperature field and the temperature gradient field in front of the molten pool based on the artificial intelligent model, delays and outputs the predicted arc energy increment to the communication module to realize the feedforward control of the arc energy, predicts the required wire feeding speed increment when the deposition process is carried out to the sampling position of the size of the previous deposition layer according to the size of the previous deposition layer, delays and outputs the predicted wire feeding speed increment to the communication module to realize the feedforward control of the wire feeding speed; the feedback controller judges the arc energy increment required by ensuring the size precision according to the real-time size of the molten pool and outputs the arc energy increment to the communication module in real time to realize the feedback control of the arc energy; the communication module is used for transmitting the control signal of the control module to the power supply to realize the real-time regulation of the arc energy, and transmitting the control signal of the control module to the wire feeder to realize the real-time regulation of the wire feeding speed; the controller self-learning module is used for training the artificial intelligence model of the feedforward controller.

Preferably, the feedforward controller is based on an artificial neuron network, the artificial neuron network predicts the required arc energy increment when the deposition process is carried out to the position through the information of the temperature field and the temperature gradient field in front of the molten pool, and predicts the required wire feeding speed increment when the deposition process is carried out to the position through the information of the size of the previous deposition layer.

Preferably, as shown in fig. 2, the controller self-learning module generates geometric models with a large variable heat dissipation condition by using a random algorithm, performs temperature field simulation by using finite element numerical simulation after layering and path planning the geometric models, obtains an arc energy increment required for ensuring size stability when a temperature field in front of a molten pool, a temperature gradient field and the molten pool move to the position during arc additive manufacturing of each geometric model, uses the data for training of an artificial intelligence model of the feedforward controller, and enables the artificial intelligence model of the feedforward controller to achieve ideal precision through iterative learning.

An adaptive control method of an arc additive manufacturing process, comprising the steps of:

step one, as shown in fig. 2, before the electric arc additive manufacturing process starts, the controller self-learning module generates different geometric models with varying heat dissipation conditions by using a random algorithm, then, the geometric models are layered and planned in a path, finite element numerical simulation of the temperature field in the additive manufacturing process is carried out, characteristic data of the temperature field in front of the molten pool and the temperature gradient field in the additive manufacturing process are extracted from simulation results, and repeatedly adjusting the arc energy to keep the size of the molten pool constant in the whole arc additive manufacturing process, and obtaining the characteristic data of the temperature field and the temperature gradient field, and the arc energy required for keeping the size constant when the molten pool moves to the position of the characteristic data sampling point form a training data set, the method is used for training the artificial intelligence model of the feedforward controller, and self-learning of the feedforward controller is carried out through a gradient descent method, so that ideal precision is achieved.

In the electric arc additive manufacturing process, the data acquisition module acquires temperature information in front of a molten pool through a front-mounted infrared thermal imager and transmits the temperature information to the control module, a front-mounted camera acquires images of the molten pool and transmits the images to the control module, and a front-mounted structured light sensor acquires contour information of a previous deposition layer and transmits the contour information to the control module;

step three, a data processor in the control module processes data transmitted by the data acquisition module in real time, calculates radiation intensity data transmitted by the infrared thermal imager to obtain temperature field information, and performs smoothing and gradient operation on the temperature field to obtain temperature gradient field information; carrying out noise reduction, binaryzation, edge detection and contour fitting on a molten pool image transmitted by a camera, and then calculating the real-time width of the molten pool; and denoising, binaryzation, edge detection, feature point extraction and contour reconstruction are carried out on the contour image of the last deposition layer transmitted by the structured light sensor, and the height and the width of the last deposition layer are calculated.

And fourthly, a feedforward controller of the control module predicts the required arc energy increment when the deposition process is carried out to the position by utilizing a trained artificial intelligence model according to the temperature and temperature gradient information of the characteristic position in front of the molten pool, delays the arc energy increment until the deposition process is carried out to the position and outputs the arc energy increment to the communication module, calculates the metal amount to be filled according to the size information of the last deposition layer, calculates the required wire feeding speed increment according to the size of the welding wire, and delays the arc energy increment until the deposition process is carried out to the position and outputs the arc energy increment to the communication module.

Calculating by a feedback controller of the control module according to the real-time size of the molten pool by using a self-adaptive PID control algorithm to obtain an arc energy increment required by ensuring the size precision, and outputting the arc energy increment to the communication module in real time;

and step six, the communication module transmits the control signal of the control module to the power supply through the power supply communication interface to realize real-time adjustment of the arc energy, and transmits the control signal of the control module to the wire feeder through the wire feeder communication interface to realize real-time adjustment of the wire feeding speed.

The invention fundamentally solves the problems of uneven forming and larger deviation of the deposition size and the design size caused by variable heat dissipation conditions in the process of the electric arc additive manufacturing of the complex structure, and improves the applicability of the electric arc additive manufacturing technology to the complex structure.

The above description of the embodiments is only intended to facilitate the understanding of the method of the invention and its core ideas; meanwhile, for a person skilled in the art, according to the idea of the present invention, there may be variations in the specific embodiments and the application scope, and in summary, the content of the present specification should not be construed as a limitation to the present invention.