CN118106976A - Control method and system of manipulator for injection molding machine - Google Patents
Control method and system of manipulator for injection molding machine Download PDFInfo
- Publication number
- CN118106976A CN118106976A CN202410532539.5A CN202410532539A CN118106976A CN 118106976 A CN118106976 A CN 118106976A CN 202410532539 A CN202410532539 A CN 202410532539A CN 118106976 A CN118106976 A CN 118106976A
- Authority
- CN
- China
- Prior art keywords
- injection molding
- molding machine
- manipulator
- data
- lstm
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000001746 injection moulding Methods 0.000 title claims abstract description 467
- 238000000034 method Methods 0.000 title claims abstract description 33
- 238000012549 training Methods 0.000 claims abstract description 42
- 239000000463 material Substances 0.000 claims abstract description 35
- 238000012360 testing method Methods 0.000 claims abstract description 32
- 239000002245 particle Substances 0.000 claims abstract description 22
- 238000012545 processing Methods 0.000 claims abstract description 14
- 230000002457 bidirectional effect Effects 0.000 claims abstract description 10
- 230000004927 fusion Effects 0.000 claims description 16
- 238000012706 support-vector machine Methods 0.000 claims description 12
- 238000012544 monitoring process Methods 0.000 claims description 10
- 239000012778 molding material Substances 0.000 claims description 9
- 230000006870 function Effects 0.000 claims description 8
- 230000008859 change Effects 0.000 claims description 6
- 238000013527 convolutional neural network Methods 0.000 claims description 6
- 238000005457 optimization Methods 0.000 claims description 6
- ORILYTVJVMAKLC-UHFFFAOYSA-N Adamantane Natural products C1C(C2)CC3CC1CC2C3 ORILYTVJVMAKLC-UHFFFAOYSA-N 0.000 claims description 4
- 230000001133 acceleration Effects 0.000 claims description 4
- 238000002347 injection Methods 0.000 claims description 4
- 239000007924 injection Substances 0.000 claims description 4
- 230000008569 process Effects 0.000 claims description 3
- 238000004519 manufacturing process Methods 0.000 description 7
- 230000008901 benefit Effects 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 4
- 238000013473 artificial intelligence Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 238000011161 development Methods 0.000 description 1
- 238000007599 discharging Methods 0.000 description 1
- 230000009191 jumping Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000010137 moulding (plastic) Methods 0.000 description 1
- 239000000243 solution Substances 0.000 description 1
Landscapes
- Injection Moulding Of Plastics Or The Like (AREA)
Abstract
The invention relates to the field of manipulator control for injection molding machines, in particular to a control method and a control system for a manipulator for an injection molding machine. Carrying out data clustering processing on the injection molding machine completion state data and the material injection molding completion state data by using a k-means fuzzy clustering algorithm to obtain training injection molding machine operation data and testing injection molding machine operation data; establishing an initial Bi-LSTM injection molding machine operation prediction model based on a Bi-LSTM bidirectional long-short-term memory network, optimizing the initial Bi-LSTM injection molding machine operation prediction model by using a PSO particle swarm algorithm, inputting training injection molding machine operation data into a target Bi-LSTM injection molding machine operation prediction model for training, and obtaining the direction in which the manipulator is to be moved; generating a control instruction of a manipulator for the injection molding machine based on the direction in which the manipulator is to be moved, and controlling the manipulator for the injection molding machine in real time according to the path planning instruction; the injection molding finished product that prevents to cause because the dynamics of snatching of manipulator is too big or undersize has snatchs the vestige and drops, promotes the accuracy and the degree of accuracy of snatching of manipulator.
Description
Technical Field
The invention relates to the field of manipulator control for injection molding machines, in particular to a control method and a control system for a manipulator for an injection molding machine.
Background
As important equipment in the plastic molding industry, the production efficiency and quality of the injection molding machine have a crucial influence on the whole production flow. In order to improve the production efficiency of the injection molding machine, reduce manual operation and production cost, a manipulator for the injection molding machine is generated. The manipulator can automatically complete operations such as material taking, material discharging, stacking and the like in the injection molding process, and greatly improves the production efficiency. In the aspect of control of a manipulator for an injection molding machine, a traditional control method often adopts simple switching value control, and the mode can realize basic automatic operation, but has larger limitations in precision, speed, stability and the like. In recent years, with the development of computer technology, sensor technology and control theory, the control method of the manipulator for the injection molding machine has been continuously improved and innovated. However, the problem that the manipulator grabs easily to drop and leaves grabbing marks caused by temperature change when injection molding is completed still exists, so that the problem that how to control the manipulator according to temperature change in an injection molding machine is a technical problem to be solved at the present stage.
Disclosure of Invention
The invention aims to solve the problems, and designs a control method and a control system of a manipulator for an injection molding machine.
The technical scheme of the invention for achieving the purpose is that in the control method of the manipulator for the injection molding machine, the control method of the manipulator for the injection molding machine comprises the following steps:
acquiring injection molding operation state data in an injection molding machine, and acquiring injection molding machine completion state data and material injection molding completion state data of the injection molding machine based on the injection molding operation state data;
carrying out data clustering processing on the injection molding machine completion state data and the material injection molding completion state data by using a k-means fuzzy clustering algorithm to obtain training injection molding machine operation data and test injection molding machine operation data;
Establishing an initial Bi-LSTM injection molding machine operation prediction model based on a Bi-LSTM bidirectional long-short-term memory network, and optimizing the initial Bi-LSTM injection molding machine operation prediction model by using a PSO particle swarm algorithm to obtain a target Bi-LSTM injection molding machine operation prediction model;
Inputting the training injection molding machine operation data into the target Bi-LSTM injection molding machine operation prediction model for training, and inputting the test injection molding machine operation data into the target Bi-LSTM injection molding machine operation prediction model for testing to obtain the direction in which the manipulator is to be moved;
Generating a control instruction of a manipulator for the injection molding machine based on the direction in which the manipulator is to be moved, generating a path planning instruction of the manipulator based on the control instruction, and controlling the manipulator for the injection molding machine in real time according to the path planning instruction;
And monitoring the temperature in the injection molding machine, and if the operating temperature of the manipulator for the injection molding machine changes, generating a manipulator control force adjustment instruction, and performing target control on the manipulator for the injection molding machine based on the manipulator control force adjustment instruction.
Further, in the above control method for a manipulator for an injection molding machine, the acquiring injection molding operation state data in the injection molding machine, acquiring injection molding machine completion state data and material injection completion state data of the injection molding machine based on the injection molding operation state data, includes:
Acquiring injection molding operation state data in an injection molding machine, wherein the injection molding operation state data at least comprises injection molding machine temperature data, injection molding machine pressure data, injection molding machine position data, injection molding machine yield data, injection molding machine process data and injection molding machine quality data;
acquiring injection molding machine completion state data and material injection completion state data of an injection molding machine based on the injection molding operation state data;
The injection molding machine completion state data are injection molding machine temperature data and injection molding machine position data of the injection molding machine after injection molding of injection molding materials is completed;
The material injection molding completion state data are injection molding finished product temperature data, injection molding finished product hardness data, injection molding finished product volume data and injection molding finished product position data of the injection molding material after the injection molding of the injection molding material is completed.
Further, in the control method of a manipulator for an injection molding machine, the performing data clustering processing on the injection molding machine completion status data and the material injection molding completion status data by using a k-means fuzzy clustering algorithm to obtain training injection molding machine operation data and test injection molding machine operation data includes:
Acquiring injection molding machine completion state data and material injection molding completion state data, and initializing the injection molding machine completion state data and the material injection molding completion state data to obtain an initial injection molding machine operation data set;
Randomly selecting K initial clustering centers of the initial injection molding machine operation data to obtain clustering centers of the initial injection molding machine operation data;
setting an initial fuzzy membership U for each data point in the initial injection molding machine operation data;
Calculating the fuzzy membership U, and calculating Euclidean distance from each data point to each clustering center for each data point and each clustering center;
Calculating fuzzy membership U1 of each cluster of the data points according to the Euclidean distance and the fuzzy membership U;
calculating a new cluster center of each cluster according to the fuzzy membership U1, wherein the new cluster center is a weighted average value of all data points, and the weight is the fuzzy membership U of the data points to the cluster;
repeating the steps to obtain target injection molding machine operation data, randomly extracting 81% of data in the target injection molding machine operation data as training injection molding machine operation data, and the rest 19% of data as test injection molding machine operation data.
Further, in the control method of the manipulator for an injection molding machine, the Bi-LSTM bidirectional long-short-term memory network is used to establish an initial Bi-LSTM operation prediction model of the injection molding machine, and the PSO particle swarm algorithm is used to optimize the initial Bi-LSTM operation prediction model of the injection molding machine to obtain a target Bi-LSTM operation prediction model of the injection molding machine, including:
Establishing an initial Bi-LSTM injection molding machine operation prediction model based on a Bi-LSTM bidirectional long-short-term memory network, wherein the initial Bi-LSTM injection molding machine operation prediction model at least comprises an input layer, an embedded layer, an average pool layer and an output layer;
extracting features of the operation data of the injection molding machine through a two-dimensional convolutional neural network, and performing cascade fusion after double-flow features are obtained to obtain fused operation data of the injection molding machine;
inputting the fusion injection molding machine operation data to a transducer module to redistribute the data weight to obtain the weight injection molding machine operation data;
Inputting the operation data of the weight injection molding machine into an SVM support vector machine improved based on a PSO particle swarm algorithm for classification treatment;
and optimizing weight parameters and bias value parameters in the SVM support vector machine through a PSO particle swarm algorithm to obtain the target Bi-LSTM injection molding machine operation prediction model.
Further, in the above method for controlling a manipulator for an injection molding machine, the step of inputting the training injection molding machine operation data into the target Bi-LSTM injection molding machine operation prediction model for training, and inputting the test injection molding machine operation data into the target Bi-LSTM injection molding machine operation prediction model for testing, to obtain a direction in which the manipulator is to be moved, includes:
Inputting the training injection molding machine operation data into the target Bi-LSTM injection molding machine operation prediction model for training;
Optimizing parameters of the target Bi-LSTM injection molding machine operation prediction model by using an Adam super-parameter optimizer;
Taking the cross entropy loss function as a loss function of the running prediction model of the target Bi-LSTM injection molding machine;
Determining the optimal super parameters of the operation prediction model of the target Bi-LSTM injection molding machine by using Bayes optimization iteration to obtain the direction in which the manipulator is to be moved;
The manipulator to-be-moved direction at least comprises a manipulator moving distance, a manipulator moving direction and a manipulator moving angle.
Further, in the above control method for a manipulator for an injection molding machine, the generating a control instruction for the manipulator for an injection molding machine based on a direction in which the manipulator is to be moved, generating a path planning instruction for the manipulator based on the control instruction, and performing real-time control for the manipulator for an injection molding machine according to the path planning instruction includes:
generating a control instruction of a manipulator for the injection molding machine based on the manipulator to-be-moved direction, wherein the control instruction comprises a manipulator motion control instruction, a manipulator gesture control instruction, a manipulator speed control instruction, a manipulator force control instruction and a manipulator IO control instruction;
And generating a path planning instruction of the manipulator based on the control instruction, wherein the path planning instruction is used for defining a path starting point, a path end point, an obstacle, a path speed and an acceleration of the manipulator, and controlling the manipulator for the injection molding machine in real time according to the path planning instruction.
Further, in the above control method for a manipulator for an injection molding machine, the monitoring of the temperature in the injection molding machine, if there is a change in the operating temperature of the manipulator for an injection molding machine, generating a manipulator control force adjustment instruction, and performing target control for the manipulator for an injection molding machine based on the manipulator control force adjustment instruction, includes:
Monitoring the temperature in the injection molding machine, if the operating temperature of the manipulator for the injection molding machine is higher than the initial temperature by 12.5%, generating a manipulator control force reducing instruction, and controlling the grabbing force of the manipulator based on the manipulator control force reducing instruction;
If the operating temperature of the manipulator for the injection molding machine is lower than the initial temperature by 32%, a manipulator control force increasing instruction is generated, and the grabbing force of the manipulator is controlled based on the manipulator control force increasing instruction.
Further, the control system of the manipulator for the injection molding machine is realized, and the control system of the manipulator for the injection molding machine comprises the following modules:
The state data acquisition module is used for acquiring injection molding operation state data in the injection molding machine and acquiring injection molding machine completion state data and material injection molding completion state data of the injection molding machine based on the injection molding operation state data;
The state data processing module is used for carrying out data clustering processing on the injection molding machine completion state data and the material injection molding completion state data by using a k-means fuzzy clustering algorithm to obtain training injection molding machine operation data and test injection molding machine operation data;
The prediction model building module is used for building an initial Bi-LSTM injection molding machine operation prediction model based on a Bi-LSTM bidirectional long-short-term memory network, and optimizing the initial Bi-LSTM injection molding machine operation prediction model by using a PSO particle swarm algorithm to obtain a target Bi-LSTM injection molding machine operation prediction model;
The manipulator to-be-moved module is used for inputting the training injection molding machine operation data into the target Bi-LSTM injection molding machine operation prediction model for training, inputting the test injection molding machine operation data into the target Bi-LSTM injection molding machine operation prediction model for testing, and obtaining the manipulator to-be-moved direction;
The manipulator control module is used for generating a control instruction of the manipulator for the injection molding machine based on the direction in which the manipulator is to be moved, generating a path planning instruction of the manipulator based on the control instruction, and controlling the manipulator for the injection molding machine in real time according to the path planning instruction;
And the manipulator adjusting module is used for monitoring the temperature in the injection molding machine, generating a manipulator control force adjusting instruction if the operating temperature of the manipulator for the injection molding machine changes, and carrying out target control on the manipulator for the injection molding machine based on the manipulator control force adjusting instruction.
Further, in the control system of a manipulator for an injection molding machine, the state data processing module includes the following submodules:
the acquisition sub-module is used for acquiring the injection molding machine completion state data and the material injection molding completion state data, and initializing the injection molding machine completion state data and the material injection molding completion state data to obtain an initial injection molding machine operation data set;
The random sub-module is used for randomly selecting K initial clustering centers of the initial injection molding machine operation data to obtain clustering centers of the initial injection molding machine operation data;
The setting sub-module is used for setting an initial fuzzy membership U for each data point in the initial injection molding machine operation data;
The distance ion module is used for calculating the fuzzy membership U, and calculating Euclidean distance from each data point to the clustering center for each data point and each clustering center;
the computing sub-module is used for computing the fuzzy membership U1 of each cluster to which the data point belongs according to the Euclidean distance and the fuzzy membership U;
The weighting sub-module is used for calculating a new cluster center of each cluster according to the fuzzy membership U1, wherein the new cluster center is a weighted average value of all data points, and the weighting is the fuzzy membership U of the data points to the cluster;
and the obtained submodule is used for repeating the steps to obtain target injection molding machine operation data, randomly extracting 81% of data in the target injection molding machine operation data to obtain training injection molding machine operation data, and the rest 19% of data are test injection molding machine operation data.
Further, in the control system of a manipulator for an injection molding machine, the prediction model building module includes the following submodules:
the building sub-module is used for building an initial Bi-LSTM injection molding machine operation prediction model based on a Bi-LSTM two-way long-short-term memory network, wherein the initial Bi-LSTM injection molding machine operation prediction model at least comprises an input layer, an embedded layer, an average pool layer and an output layer;
The fusion sub-module is used for extracting the characteristics of the operation data of the injection molding machine through the two-dimensional convolutional neural network, and performing cascade fusion after obtaining double-flow characteristics to obtain the operation data of the fusion injection molding machine;
the input sub-module is used for inputting the fusion injection molding machine operation data to the transducer module to redistribute the data weight so as to obtain the weight injection molding machine operation data;
the classifying sub-module is used for inputting the running data of the weight injection molding machine into an SVM support vector machine improved based on a PSO particle swarm algorithm for classifying;
And the parameter sub-module is used for optimizing the weight parameter and the bias value parameter in the SVM support vector machine through a PSO particle swarm algorithm to obtain the target Bi-LSTM injection molding machine operation prediction model.
The automatic injection molding machine has the beneficial effects that the mechanical arm can realize automatic identification, positioning and grabbing of injection molding pieces, so that the production efficiency and accuracy are further improved. The application of the artificial intelligence technology can enable the manipulator to have learning capability, and the operation precision and efficiency of the manipulator are gradually improved through continuous learning and optimization. The temperature of the injection molding machine can be monitored, so that the grabbing force of the manipulator is adjusted according to the real-time state of the injection molding finished product, the injection molding finished product which is caused by overlarge or small grabbing force of the manipulator is prevented from having grabbing marks and falling, and the grabbing accuracy and the accuracy of the manipulator are improved.
Drawings
Various other advantages and benefits will become apparent to those of ordinary skill in the art upon reading the following detailed description of the preferred embodiments. The drawings are only for purposes of illustrating the preferred embodiments and are not to be construed as limiting the invention.
FIG. 1 is a schematic view of a first embodiment of a method for controlling a manipulator for an injection molding machine according to an embodiment of the present invention;
FIG. 2 is a schematic view showing a second embodiment of a method for controlling a manipulator for an injection molding machine according to an embodiment of the present invention;
fig. 3 is a schematic diagram of a third embodiment of a control method of a manipulator for an injection molding machine according to an embodiment of the present invention.
Fig. 4 is a schematic diagram of a first embodiment of a control system of a manipulator for an injection molding machine according to an embodiment of the present invention.
Detailed Description
The present invention will be described in further detail with reference to the drawings and examples, in order to make the objects, technical solutions and advantages of the present invention more apparent. It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the scope of the invention.
As used herein, the singular forms "a", "an" and "the" are intended to include the plural forms as well, unless expressly stated otherwise, as understood by those skilled in the art. It will be further understood that the terms "comprises" and/or "comprising," when used in this specification, specify the presence of stated features, integers, steps, operations, elements, and/or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and/or groups thereof.
The present invention will be described in detail below with reference to the accompanying drawings, as shown in fig. 1, a control method of a manipulator for an injection molding machine, the control method of the manipulator for the injection molding machine comprising the steps of:
Step 101, acquiring injection molding operation state data in an injection molding machine, and acquiring injection molding machine completion state data and material injection molding completion state data of the injection molding machine based on the injection molding operation state data;
Specifically, injection molding operation state data in the injection molding machine is obtained, wherein the injection molding operation state data at least comprises injection molding machine temperature data, injection molding machine pressure data, injection molding machine position data, injection molding machine yield data, injection molding machine process data and injection molding machine quality data; acquiring injection molding machine completion state data and material injection molding completion state data of an injection molding machine based on injection molding operation state data; the injection molding machine completion state data are injection molding machine temperature data and injection molding machine position data of the injection molding machine after injection molding of the injection molding material is completed; the material injection molding completion state data are injection molding finished product temperature data, injection molding finished product hardness data, injection molding finished product volume data and injection molding finished product position data of the injection molding material after the injection molding of the injection molding material is completed.
102, Performing data clustering processing on the injection molding machine completion state data and the material injection molding completion state data by using a k-means fuzzy clustering algorithm to obtain training injection molding machine operation data and test injection molding machine operation data;
Specifically, in this embodiment, the completion status data of the injection molding machine and the material injection molding completion status data are obtained, and the completion status data of the injection molding machine and the material injection molding completion status data are data initialized to obtain an initial operation data set of the injection molding machine; randomly selecting K initial clustering centers of the initial injection molding machine operation data to obtain clustering centers of the initial injection molding machine operation data; setting an initial fuzzy membership U for each data point in initial injection molding machine operation data; calculating fuzzy membership U, and calculating Euclidean distance from each data point to each clustering center for each data point and each clustering center; calculating fuzzy membership U1 of each cluster of the data points according to the Euclidean distance and the fuzzy membership U; calculating a new cluster center of each cluster according to the fuzzy membership U1, wherein the new cluster center is a weighted average value of all data points, and the weight is the fuzzy membership U of the data points to the cluster; repeating the steps to obtain target injection molding machine operation data, randomly extracting 81% of data in the target injection molding machine operation data as training injection molding machine operation data, and the rest 19% of data as test injection molding machine operation data.
Step 103, establishing an initial Bi-LSTM injection molding machine operation prediction model based on a Bi-LSTM two-way long-short-term memory network, and optimizing the initial Bi-LSTM injection molding machine operation prediction model by using a PSO particle swarm algorithm to obtain a target Bi-LSTM injection molding machine operation prediction model;
Specifically, in this embodiment, an initial Bi-LSTM injection molding machine operation prediction model is established based on a Bi-LSTM bidirectional long-short-term memory network, where the initial Bi-LSTM injection molding machine operation prediction model includes at least an input layer, an embedded layer, an average pool layer, and an output layer; extracting features of the operation data of the injection molding machine through a two-dimensional convolutional neural network, and performing cascade fusion after double-flow features are obtained to obtain fused operation data of the injection molding machine; inputting the fusion injection molding machine operation data into a transducer module to redistribute the data weight so as to obtain the weight injection molding machine operation data; inputting the running data of the weight injection molding machine into an SVM support vector machine improved based on a PSO particle swarm algorithm for classification treatment; and optimizing weight parameters and bias value parameters in the SVM support vector machine through a PSO particle swarm algorithm to obtain the target Bi-LSTM injection molding machine operation prediction model.
104, Inputting training injection molding machine operation data into a target Bi-LSTM injection molding machine operation prediction model for training, and inputting test injection molding machine operation data into the target Bi-LSTM injection molding machine operation prediction model for testing to obtain the direction in which the manipulator is to be moved;
Specifically, in this embodiment, the operation data of the training injection molding machine is input into the operation prediction model of the target Bi-LSTM injection molding machine for training; optimizing parameters of a target Bi-LSTM injection molding machine operation prediction model by using an Adam super-parameter optimizer; taking the cross entropy loss function as a loss function of a target Bi-LSTM injection molding machine operation prediction model; determining the optimal super parameters of the operation prediction model of the target Bi-LSTM injection molding machine by using Bayes optimization iteration to obtain the direction in which the manipulator is to be moved; the manipulator to-be-moved direction at least comprises a manipulator moving distance, a manipulator moving direction and a manipulator moving angle.
Step 105, generating a control instruction of the manipulator for the injection molding machine based on the direction in which the manipulator is to be moved, generating a path planning instruction of the manipulator based on the control instruction, and controlling the manipulator for the injection molding machine in real time according to the path planning instruction;
Specifically, in this embodiment, a control instruction of a manipulator for an injection molding machine is generated based on a direction in which the manipulator is to be moved, where the control instruction includes a manipulator motion control instruction, a manipulator posture control instruction, a manipulator speed control instruction, a manipulator force control instruction, and a manipulator IO control instruction; and generating a path planning instruction of the manipulator based on the control instruction, wherein the path planning instruction is used for defining a path starting point, a path end point, an obstacle, a path speed and an acceleration of the manipulator, and controlling the manipulator for the injection molding machine in real time according to the path planning instruction.
Specifically, the motion control instruction in this embodiment includes:
MOVJ: for articulation, i.e. angular movement, of the manipulator. The manipulator can change the joint angle according to a certain speed and acceleration. MOVL: the linear motion is used for enabling the manipulator to move along a linear track. PTP (Point-to-Point): for point-to-point movement, to move the robot from one position to another.
The gesture control instruction includes: PULSE: is used for controlling the gesture changes such as rotation or overturn of the manipulator. PTP (in attitude control): the same can be used for point-to-point control of gestures.
The speed control instruction includes: JOG (JOG): the device is used for realizing rapid adjustment of the movement speed of the manipulator. LIN (Linear): the device is used for realizing uniform movement of the manipulator.
The force control instructions include: FORCE: so that the manipulator can maintain certain strength or pressure. POS: the manipulator is kept at a certain position or posture.
The IO control instruction includes: SETDO: for setting the state of a digital output port, such as controlling a pneumatic valve, etc. WAITDI: for awaiting a change in the state of the digital input port, such as awaiting a signal from a sensor.
Other control instructions include: get: for reading data from the specified register address. Put: for storing data into a specified register address. Jump: for jumping to a specified program address.
And 106, monitoring the temperature in the injection molding machine, and if the operating temperature of the manipulator for the injection molding machine changes, generating a manipulator control force adjustment instruction, and performing target control on the manipulator for the injection molding machine based on the manipulator control force adjustment instruction.
Specifically, in this embodiment, the temperature in the injection molding machine is monitored, if the operating temperature of the manipulator for the injection molding machine is higher than the initial temperature by 12.5%, a manipulator control force reducing instruction is generated, and the grabbing force of the manipulator is controlled based on the manipulator control force reducing instruction; if the operating temperature of the manipulator for the injection molding machine is lower than the initial temperature by 32%, a manipulator control force increasing instruction is generated, and the grabbing force of the manipulator is controlled based on the manipulator control force increasing instruction.
The automatic injection molding machine has the beneficial effects that the mechanical arm can realize automatic identification, positioning and grabbing of injection molding pieces, so that the production efficiency and accuracy are further improved. The application of the artificial intelligence technology can enable the manipulator to have learning capability, and the operation precision and efficiency of the manipulator are gradually improved through continuous learning and optimization. The temperature of the injection molding machine can be monitored, so that the grabbing force of the manipulator is adjusted according to the real-time state of the injection molding finished product, the injection molding finished product which is caused by overlarge or small grabbing force of the manipulator is prevented from having grabbing marks and falling, and the grabbing accuracy and the accuracy of the manipulator are improved.
In this embodiment, referring to fig. 2, in a second embodiment of a control method for a manipulator for an injection molding machine according to the present invention, an initial Bi-LSTM operation prediction model is established based on a Bi-LSTM two-way long-short-term memory network, and the initial Bi-LSTM operation prediction model is optimized by using a PSO particle swarm algorithm to obtain a target Bi-LSTM operation prediction model, including:
Step 201, establishing an initial Bi-LSTM injection molding machine operation prediction model based on a Bi-LSTM two-way long-short-term memory network, wherein the initial Bi-LSTM injection molding machine operation prediction model at least comprises an input layer, an embedded layer, an average pool layer and an output layer;
step 202, extracting features of operation data of the injection molding machine through a two-dimensional convolutional neural network, and performing cascade fusion after double-flow features are obtained to obtain fused operation data of the injection molding machine;
Step 203, inputting the operation data of the fusion injection molding machine to a transducer module to redistribute the data weight so as to obtain the operation data of the weight injection molding machine;
Step 204, inputting the running data of the weight injection molding machine into an SVM support vector machine improved based on a PSO particle swarm algorithm for classification treatment;
And 205, optimizing weight parameters and bias value parameters in the SVM support vector machine through a PSO particle swarm algorithm to obtain a target Bi-LSTM injection molding machine operation prediction model.
In this embodiment, referring to fig. 3, in a third embodiment of a control method for a manipulator for an injection molding machine according to the present invention, training injection molding machine operation data is input into a target Bi-LSTM injection molding machine operation prediction model for training, test injection molding machine operation data is input into the target Bi-LSTM injection molding machine operation prediction model for testing, and a to-be-moved direction of the manipulator is obtained, including:
Step 301, inputting the operation data of the training injection molding machine into a target Bi-LSTM injection molding machine operation prediction model for training;
Step 302, optimizing parameters of a target Bi-LSTM injection molding machine operation prediction model by using an Adam super-parameter optimizer;
step 303, using the cross entropy loss function as a loss function of a target Bi-LSTM injection molding machine operation prediction model;
Step 304, determining the optimal super parameters of the operation prediction model of the target Bi-LSTM injection molding machine by using Bayes optimization iteration to obtain the direction in which the manipulator is to be moved;
in step 305, the manipulator to-be-moved direction at least includes a manipulator movement distance, a manipulator movement direction and a manipulator movement angle.
The control method of the manipulator for an injection molding machine provided by the embodiment of the present invention is described above, and the control system of the manipulator for an injection molding machine according to the embodiment of the present invention is described below, referring to fig. 4, where an embodiment of the control system of the embodiment of the present invention includes:
The state data acquisition module is used for acquiring injection molding operation state data in the injection molding machine and acquiring injection molding machine completion state data and material injection molding completion state data of the injection molding machine based on the injection molding operation state data;
The state data processing module is used for carrying out data clustering processing on the injection molding machine completion state data and the material injection molding completion state data by using a k-means fuzzy clustering algorithm to obtain training injection molding machine operation data and test injection molding machine operation data;
The prediction model building module is used for building an initial Bi-LSTM injection molding machine operation prediction model based on a Bi-LSTM bidirectional long-short-term memory network, optimizing the initial Bi-LSTM injection molding machine operation prediction model by using a PSO particle swarm algorithm, and obtaining a target Bi-LSTM injection molding machine operation prediction model;
the manipulator to-be-moved module is used for inputting the operation data of the training injection molding machine into the operation prediction model of the target Bi-LSTM injection molding machine for training, and inputting the operation data of the test injection molding machine into the operation prediction model of the target Bi-LSTM injection molding machine for testing, so as to obtain the to-be-moved direction of the manipulator;
The manipulator control module is used for generating a control instruction of the manipulator for the injection molding machine based on the direction in which the manipulator is to be moved, generating a path planning instruction of the manipulator based on the control instruction, and controlling the manipulator for the injection molding machine in real time according to the path planning instruction;
And the manipulator adjusting module is used for monitoring the temperature in the injection molding machine, generating a manipulator control force adjusting instruction if the operating temperature of the manipulator for the injection molding machine changes, and performing target control on the manipulator for the injection molding machine based on the manipulator control force adjusting instruction.
The foregoing has shown and described the basic principles, principal features and advantages of the invention. It will be understood by those skilled in the art that the present invention is not limited to the above-described embodiments, and that the above-described embodiments and descriptions are only preferred embodiments of the present invention, and are not intended to limit the invention, and that various changes and modifications may be made therein without departing from the spirit and scope of the invention as claimed. The scope of the invention is defined by the appended claims and equivalents thereof.
Claims (10)
1. The control method of the manipulator for the injection molding machine is characterized by comprising the following steps of:
acquiring injection molding operation state data in an injection molding machine, and acquiring injection molding machine completion state data and material injection molding completion state data of the injection molding machine based on the injection molding operation state data;
carrying out data clustering processing on the injection molding machine completion state data and the material injection molding completion state data by using a k-means fuzzy clustering algorithm to obtain training injection molding machine operation data and test injection molding machine operation data;
Establishing an initial Bi-LSTM injection molding machine operation prediction model based on a Bi-LSTM bidirectional long-short-term memory network, and optimizing the initial Bi-LSTM injection molding machine operation prediction model by using a PSO particle swarm algorithm to obtain a target Bi-LSTM injection molding machine operation prediction model;
Inputting the training injection molding machine operation data into the target Bi-LSTM injection molding machine operation prediction model for training, and inputting the test injection molding machine operation data into the target Bi-LSTM injection molding machine operation prediction model for testing to obtain the direction in which the manipulator is to be moved;
Generating a control instruction of a manipulator for the injection molding machine based on the direction in which the manipulator is to be moved, generating a path planning instruction of the manipulator based on the control instruction, and controlling the manipulator for the injection molding machine in real time according to the path planning instruction;
And monitoring the temperature in the injection molding machine, and if the operating temperature of the manipulator for the injection molding machine changes, generating a manipulator control force adjustment instruction, and performing target control on the manipulator for the injection molding machine based on the manipulator control force adjustment instruction.
2. The method for controlling a manipulator for an injection molding machine according to claim 1, wherein the acquiring injection molding operation state data in the injection molding machine, acquiring injection molding machine completion state data and material injection completion state data of the injection molding machine based on the injection molding operation state data, comprises:
Acquiring injection molding operation state data in an injection molding machine, wherein the injection molding operation state data at least comprises injection molding machine temperature data, injection molding machine pressure data, injection molding machine position data, injection molding machine yield data, injection molding machine process data and injection molding machine quality data;
acquiring injection molding machine completion state data and material injection completion state data of an injection molding machine based on the injection molding operation state data;
The injection molding machine completion state data are injection molding machine temperature data and injection molding machine position data of the injection molding machine after injection molding of injection molding materials is completed;
The material injection molding completion state data are injection molding finished product temperature data, injection molding finished product hardness data, injection molding finished product volume data and injection molding finished product position data of the injection molding material after the injection molding of the injection molding material is completed.
3. The method for controlling a manipulator for an injection molding machine according to claim 1, wherein the performing data clustering processing on the injection molding machine completion status data and the material injection molding completion status data by using a k-means fuzzy clustering algorithm to obtain training injection molding machine operation data and test injection molding machine operation data comprises:
Acquiring injection molding machine completion state data and material injection molding completion state data, and initializing the injection molding machine completion state data and the material injection molding completion state data to obtain an initial injection molding machine operation data set;
Randomly selecting K initial clustering centers of the initial injection molding machine operation data to obtain clustering centers of the initial injection molding machine operation data;
setting an initial fuzzy membership U for each data point in the initial injection molding machine operation data;
Calculating the fuzzy membership U, and calculating Euclidean distance from each data point to each clustering center for each data point and each clustering center;
Calculating fuzzy membership U1 of each cluster of the data points according to the Euclidean distance and the fuzzy membership U;
calculating a new cluster center of each cluster according to the fuzzy membership U1, wherein the new cluster center is a weighted average value of all data points, and the weight is the fuzzy membership U of the data points to the cluster;
repeating the steps to obtain target injection molding machine operation data, randomly extracting 81% of data in the target injection molding machine operation data as training injection molding machine operation data, and the rest 19% of data as test injection molding machine operation data.
4. The control method of a manipulator for an injection molding machine according to claim 1, wherein the creating an initial Bi-LSTM injection molding machine operation prediction model based on a Bi-LSTM two-way long-short-term memory network, optimizing the initial Bi-LSTM injection molding machine operation prediction model by using a PSO particle swarm algorithm, and obtaining a target Bi-LSTM injection molding machine operation prediction model, includes:
Establishing an initial Bi-LSTM injection molding machine operation prediction model based on a Bi-LSTM bidirectional long-short-term memory network, wherein the initial Bi-LSTM injection molding machine operation prediction model at least comprises an input layer, an embedded layer, an average pool layer and an output layer;
extracting features of the operation data of the injection molding machine through a two-dimensional convolutional neural network, and performing cascade fusion after double-flow features are obtained to obtain fused operation data of the injection molding machine;
inputting the fusion injection molding machine operation data to a transducer module to redistribute the data weight to obtain the weight injection molding machine operation data;
Inputting the operation data of the weight injection molding machine into an SVM support vector machine improved based on a PSO particle swarm algorithm for classification treatment;
and optimizing weight parameters and bias value parameters in the SVM support vector machine through a PSO particle swarm algorithm to obtain the target Bi-LSTM injection molding machine operation prediction model.
5. The method for controlling a manipulator for an injection molding machine according to claim 1, wherein the step of inputting the training injection molding machine operation data into the target Bi-LSTM injection molding machine operation prediction model for training, and inputting the test injection molding machine operation data into the target Bi-LSTM injection molding machine operation prediction model for testing, to obtain the direction in which the manipulator is to be moved, comprises:
Inputting the training injection molding machine operation data into the target Bi-LSTM injection molding machine operation prediction model for training;
Optimizing parameters of the target Bi-LSTM injection molding machine operation prediction model by using an Adam super-parameter optimizer;
Taking the cross entropy loss function as a loss function of the running prediction model of the target Bi-LSTM injection molding machine;
Determining the optimal super parameters of the operation prediction model of the target Bi-LSTM injection molding machine by using Bayes optimization iteration to obtain the direction in which the manipulator is to be moved;
The manipulator to-be-moved direction at least comprises a manipulator moving distance, a manipulator moving direction and a manipulator moving angle.
6. The method for controlling a manipulator for an injection molding machine according to claim 1, wherein generating a control command for the manipulator for an injection molding machine based on a direction in which the manipulator is to be moved, generating a path planning command for the manipulator based on the control command, and controlling the manipulator for an injection molding machine in real time based on the path planning command, comprises:
generating a control instruction of a manipulator for the injection molding machine based on the manipulator to-be-moved direction, wherein the control instruction comprises a manipulator motion control instruction, a manipulator gesture control instruction, a manipulator speed control instruction, a manipulator force control instruction and a manipulator IO control instruction;
And generating a path planning instruction of the manipulator based on the control instruction, wherein the path planning instruction is used for defining a path starting point, a path end point, an obstacle, a path speed and an acceleration of the manipulator, and controlling the manipulator for the injection molding machine in real time according to the path planning instruction.
7. The method according to claim 1, wherein the monitoring of the temperature in the injection molding machine, if there is a change in the operating temperature of the injection molding machine manipulator, generating a manipulator control force adjustment command, and performing target control of the injection molding machine manipulator based on the manipulator control force adjustment command, comprises:
Monitoring the temperature in the injection molding machine, if the operating temperature of the manipulator for the injection molding machine is higher than the initial temperature by 12.5%, generating a manipulator control force reducing instruction, and controlling the grabbing force of the manipulator based on the manipulator control force reducing instruction;
If the operating temperature of the manipulator for the injection molding machine is lower than the initial temperature by 32%, a manipulator control force increasing instruction is generated, and the grabbing force of the manipulator is controlled based on the manipulator control force increasing instruction.
8. The control system of the manipulator for the injection molding machine is characterized by comprising the following modules:
The state data acquisition module is used for acquiring injection molding operation state data in the injection molding machine and acquiring injection molding machine completion state data and material injection molding completion state data of the injection molding machine based on the injection molding operation state data;
The state data processing module is used for carrying out data clustering processing on the injection molding machine completion state data and the material injection molding completion state data by using a k-means fuzzy clustering algorithm to obtain training injection molding machine operation data and test injection molding machine operation data;
The prediction model building module is used for building an initial Bi-LSTM injection molding machine operation prediction model based on a Bi-LSTM bidirectional long-short-term memory network, and optimizing the initial Bi-LSTM injection molding machine operation prediction model by using a PSO particle swarm algorithm to obtain a target Bi-LSTM injection molding machine operation prediction model;
The manipulator to-be-moved module is used for inputting the training injection molding machine operation data into the target Bi-LSTM injection molding machine operation prediction model for training, inputting the test injection molding machine operation data into the target Bi-LSTM injection molding machine operation prediction model for testing, and obtaining the manipulator to-be-moved direction;
The manipulator control module is used for generating a control instruction of the manipulator for the injection molding machine based on the direction in which the manipulator is to be moved, generating a path planning instruction of the manipulator based on the control instruction, and controlling the manipulator for the injection molding machine in real time according to the path planning instruction;
And the manipulator adjusting module is used for monitoring the temperature in the injection molding machine, generating a manipulator control force adjusting instruction if the operating temperature of the manipulator for the injection molding machine changes, and carrying out target control on the manipulator for the injection molding machine based on the manipulator control force adjusting instruction.
9. The control system of a manipulator for an injection molding machine of claim 8, wherein the status data processing module includes the following sub-modules:
the acquisition sub-module is used for acquiring the injection molding machine completion state data and the material injection molding completion state data, and initializing the injection molding machine completion state data and the material injection molding completion state data to obtain an initial injection molding machine operation data set;
The random sub-module is used for randomly selecting K initial clustering centers of the initial injection molding machine operation data to obtain clustering centers of the initial injection molding machine operation data;
The setting sub-module is used for setting an initial fuzzy membership U for each data point in the initial injection molding machine operation data;
The distance ion module is used for calculating the fuzzy membership U, and calculating Euclidean distance from each data point to the clustering center for each data point and each clustering center;
the computing sub-module is used for computing the fuzzy membership U1 of each cluster to which the data point belongs according to the Euclidean distance and the fuzzy membership U;
The weighting sub-module is used for calculating a new cluster center of each cluster according to the fuzzy membership U1, wherein the new cluster center is a weighted average value of all data points, and the weighting is the fuzzy membership U of the data points to the cluster;
and the obtained submodule is used for repeating the steps to obtain target injection molding machine operation data, randomly extracting 81% of data in the target injection molding machine operation data to obtain training injection molding machine operation data, and the rest 19% of data are test injection molding machine operation data.
10. The control system of a manipulator for an injection molding machine according to claim 8, wherein the prediction model creation module includes the following submodules:
the building sub-module is used for building an initial Bi-LSTM injection molding machine operation prediction model based on a Bi-LSTM two-way long-short-term memory network, wherein the initial Bi-LSTM injection molding machine operation prediction model at least comprises an input layer, an embedded layer, an average pool layer and an output layer;
The fusion sub-module is used for extracting the characteristics of the operation data of the injection molding machine through the two-dimensional convolutional neural network, and performing cascade fusion after obtaining double-flow characteristics to obtain the operation data of the fusion injection molding machine;
the input sub-module is used for inputting the fusion injection molding machine operation data to the transducer module to redistribute the data weight so as to obtain the weight injection molding machine operation data;
the classifying sub-module is used for inputting the running data of the weight injection molding machine into an SVM support vector machine improved based on a PSO particle swarm algorithm for classifying;
And the parameter sub-module is used for optimizing the weight parameter and the bias value parameter in the SVM support vector machine through a PSO particle swarm algorithm to obtain the target Bi-LSTM injection molding machine operation prediction model.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410532539.5A CN118106976B (en) | 2024-04-30 | 2024-04-30 | Control method and system of manipulator for injection molding machine |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410532539.5A CN118106976B (en) | 2024-04-30 | 2024-04-30 | Control method and system of manipulator for injection molding machine |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN118106976A true CN118106976A (en) | 2024-05-31 |
| CN118106976B CN118106976B (en) | 2024-06-28 |
Family
ID=91209056
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410532539.5A Active CN118106976B (en) | 2024-04-30 | 2024-04-30 | Control method and system of manipulator for injection molding machine |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN118106976B (en) |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP3227061A1 (en) * | 2014-12-05 | 2017-10-11 | Broetje-Automation GmbH | Method for motion simulation of a manipulator |
| CN109571898A (en) * | 2018-11-28 | 2019-04-05 | 宁波安信数控技术有限公司 | A kind of manipulator of injection machine precision compensation system and method |
| CN110625899A (en) * | 2019-10-31 | 2019-12-31 | 苏州超群智能科技有限公司 | Manipulator for injection molding machine capable of automatically measuring temperature in mold |
| KR20210012672A (en) * | 2019-07-26 | 2021-02-03 | 한국생산기술연구원 | System and method for automatic control of robot manipulator based on artificial intelligence |
| CN112405543A (en) * | 2020-11-23 | 2021-02-26 | 长沙理工大学 | Mechanical arm dense object temperature-first grabbing method based on deep reinforcement learning |
| CN112906155A (en) * | 2021-02-05 | 2021-06-04 | 深圳市浦联智能科技有限公司 | Virtual measurement method for injection molding product information |
| CN117348573A (en) * | 2023-11-20 | 2024-01-05 | 峰睿新科智能有限公司 | Method and system for controlling injection molding production quality in real time |
-
2024
- 2024-04-30 CN CN202410532539.5A patent/CN118106976B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP3227061A1 (en) * | 2014-12-05 | 2017-10-11 | Broetje-Automation GmbH | Method for motion simulation of a manipulator |
| CN109571898A (en) * | 2018-11-28 | 2019-04-05 | 宁波安信数控技术有限公司 | A kind of manipulator of injection machine precision compensation system and method |
| KR20210012672A (en) * | 2019-07-26 | 2021-02-03 | 한국생산기술연구원 | System and method for automatic control of robot manipulator based on artificial intelligence |
| CN110625899A (en) * | 2019-10-31 | 2019-12-31 | 苏州超群智能科技有限公司 | Manipulator for injection molding machine capable of automatically measuring temperature in mold |
| CN112405543A (en) * | 2020-11-23 | 2021-02-26 | 长沙理工大学 | Mechanical arm dense object temperature-first grabbing method based on deep reinforcement learning |
| CN112906155A (en) * | 2021-02-05 | 2021-06-04 | 深圳市浦联智能科技有限公司 | Virtual measurement method for injection molding product information |
| CN117348573A (en) * | 2023-11-20 | 2024-01-05 | 峰睿新科智能有限公司 | Method and system for controlling injection molding production quality in real time |
Also Published As
| Publication number | Publication date |
|---|---|
| CN118106976B (en) | 2024-06-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110450156B (en) | Optimal design method of self-adaptive fuzzy controller of multi-degree-of-freedom mechanical arm system | |
| CN110450155B (en) | Optimal design method for controller of multi-degree-of-freedom mechanical arm system | |
| Jiang et al. | A review of robotic assembly strategies for the full operation procedure: planning, execution and evaluation | |
| CN111872934A (en) | Mechanical arm control method and system based on hidden semi-Markov model | |
| Minami et al. | Manipulator visual servoing and tracking of fish using a genetic algorithm | |
| US10481565B2 (en) | Methods and systems for nonlinear adaptive control and filtering | |
| US11691278B2 (en) | Hybrid computing achitectures with specialized processors to encode/decode latent representations for controlling dynamic mechanical systems | |
| US20230109398A1 (en) | Expedited robot teach-through initialization from previously trained system | |
| Chu et al. | Robotic grasp detection using a novel two-stage approach | |
| Aljalbout et al. | Learning vision-based reactive policies for obstacle avoidance | |
| Suzuki et al. | Online self-supervised learning for object picking: detecting optimum grasping position using a metric learning approach | |
| CN115781685A (en) | High-precision mechanical arm control method and system based on reinforcement learning | |
| Zhao et al. | Model accelerated reinforcement learning for high precision robotic assembly | |
| CN118106976B (en) | Control method and system of manipulator for injection molding machine | |
| Liu et al. | Neural network control system of cooperative robot based on genetic algorithms | |
| CN113910218B (en) | Robot calibration method and device based on kinematic and deep neural network fusion | |
| Sheng et al. | Forward kinematics of the Stewart platform using hybrid immune genetic algorithm | |
| CN112388620B (en) | Trajectory tracking control algorithm for pneumatic muscle driving system | |
| Wang et al. | Deep Reinforcement Learning for Peg-in-hole Assembly Task Via Information Utilization Method | |
| Udawatta et al. | Control of underactuated robot manipulators using switching computed torque method: GA based approach | |
| CN115344047A (en) | Robot switching type predictive control trajectory tracking method based on neural network model | |
| Keymeulen et al. | Comparison between an off-line model-free and an on-line model-based evolution applied to a robotics navigation system using evolvable hardware | |
| Ou et al. | Support vector machine based approach for abstracting human control strategy in controlling dynamically stable robots | |
| Zhang et al. | Control method of shaft and hole mating based on convolution neural network in assembly building prefabricated components | |
| Chan et al. | Evolutionary computation for on-line and off-line parameter tuning of evolving fuzzy neural networks |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |