Industrial Robot Control System and Method Based on Deep Learning

Abstract

The present invention discloses an industrial robot control method based on deep learning, which relates to the technical field of industrial robot control; it includes collecting the operation data of the robot in real time through sensors; dynamically adjusting control parameters according to the changes in real-time data; based on a reinforcement learning model, analyzing the complexity of the current task and the capabilities of the robot in real time, and intelligently allocating the collaborative tasks among multiple robots; continuously monitoring the operation state of the robot and issuing early warnings when abnormalities are found; through the reinforcement learning model, the present invention analyzes and intelligently allocates the task complexity and the capabilities of the robot in real time, optimizes the task collaboration efficiency and improves the resource utilization rate; constructs an anomaly detection and early warning mechanism, quickly identifies the running deviation state, feeds back to the control system through a hierarchical early warning signal when an anomaly occurs, and reallocates tasks or adjusts parameters, significantly improving the adaptability, collaboration ability and operation safety of industrial robots in complex scenarios.

Description

Technical Field

[0001] The present invention relates to the technical field of industrial robot control, and particularly to an industrial robot control system and method based on deep learning. Background Art

[0002] In the field of industrial robot control, traditional control methods mainly rely on preset motion trajectories and fixed control parameters to achieve basic mechanical actions; these methods perform well in single and repetitive working environments, but when the task environment is complex and variable or involves multi-robot collaboration, the limitations of traditional methods gradually emerge; with the increasing diversification of industrial robot application scenarios and the continuous improvement of task complexity, simple preset parameters and trajectories are difficult to meet actual needs; especially in the face of real-time data changes, how to dynamically adjust control parameters to adapt to the real-time environment has become one of the important technical challenges in robot control.

[0003] However, traditional control methods are difficult to effectively evaluate the complexity of tasks and the capabilities of robots themselves, resulting in unreasonable multi-robot collaboration task allocation and low resource utilization efficiency; in addition, in the face of abnormal operating states (such as overloaded load, abnormal speed or abnormal energy consumption), traditional methods cannot detect and issue warning signals in time, which not only increases the risk of system operation but may also cause irreversible damage to the equipment. Summary of the Invention

[0004] In view of the problems existing in the existing industrial robot control technology, the present invention proposes an industrial robot control system and method based on deep learning.

[0005] Therefore, the problem to be solved by the present invention is how to solve the intelligent task allocation, dynamic anomaly detection and warning problems of multi-robot collaboration.

[0006] To solve the above technical problems, the present invention provides the following technical solutions:

[0007] In a first aspect, an embodiment of the present invention provides an industrial robot control method based on deep learning, which includes collecting real-time operation data of the robot during work through sensors, including position, speed and load information; dynamically adjusting control parameters according to real-time data changes; based on a reinforcement learning model, analyzing the complexity of the current task and the capabilities of the robot in real time, and intelligently allocating collaboration tasks among multiple robots; continuously monitoring the running state of the robot and issuing a warning when an anomaly is detected.

[0008] As a preferred embodiment of the industrial robot control method based on deep learning according to the present invention, the following steps are included: Install a variety of sensors on the industrial robot, including position sensors, speed sensors, torque sensors, and environmental sensors, to comprehensively monitor the operating state of the robot and the working environment; Regularly collect the real-time data of the robot during task execution, and the real-time data includes position data, speed data, torque data, and environmental data.

[0009] Store the collected data in a database and perform preprocessing, cleaning, and standardization of the data for subsequent analysis.

[0010] As a preferred embodiment of the industrial robot control method based on deep learning according to the present invention, the following steps are included: Extract features from the real-time data, and the features include the current load , the current speed , the current environmental temperature , and the workpiece characteristic parameters , and the workpiece characteristic parameters include material parameters and shape parameters.

[0011] According to the extracted features, establish a dynamic parameter model, map the real-time data to the adjustment range of control parameters, and define key control parameters, including speed gain, acceleration limit, and load response time, as follows:

[0012] Based on the features, establish a mapping relationship, convert the extracted features into the adjustment range of control parameters, and set control parameters, including speed gain , acceleration limit , and load response time , and the formulas for the speed gain , acceleration limit , and load response time are respectively:

[0013] ;

[0014] ;

[0015] ;

[0016] where, represents the reference speed gain, represents the workpiece characteristic coefficient, represents the environmental temperature coefficient, represents the workpiece characteristic reference value, represents the environmental temperature reference value, represents the attenuation factor for controlling the speed, represents the reference acceleration limit, Represents the load impact coefficient, Represents the environmental temperature impact coefficient, Represents the load upper limit reference value, Represents the speed impact coefficient, Represents the speed upper limit reference value, Represents the reference response time, Represents the load impact coefficient, Represents the environmental temperature impact coefficient, Represents the load reference value.

[0017] As a preferred embodiment of the industrial robot control method based on deep learning according to the present invention, wherein: applying the dynamic parameter model to the control strategy, specifically as follows:

[0018] Using the control parameters generated by the dynamic parameter model, dynamically adjust during the operation of the robot, and adjust the speed gain according to the real-time monitoring data and the acceleration limit , and the adjustment formula is as follows:

[0019] ;

[0020] ;

[0021] Wherein, Represents the safety threshold of the preset load, and Represents the preset adjustment coefficient.

[0022] Introduce a performance feedback mechanism, dynamically update the feature extraction and parameter model by evaluating the actual performance of the robot under different conditions, and achieve the self-learning ability based on historical data.

[0023] As a preferred embodiment of the industrial robot control method based on deep learning according to the present invention, wherein: by analyzing the attributes of the current task, extract the task complexity index.

[0024] Real-time monitor the working status and performance of each robot, including the current load, energy consumption and remaining task capabilities, and form a robot ability evaluation model.

[0025] Based on the evaluation results of task complexity and robot ability, construct a reinforcement learning decision model, train the model through historical data, and optimize the task allocation strategy. The specific process is as follows:

[0026] Collect and label the features and historical allocation data of each task, including task complexity parameters and robot ability parameters. The task complexity parameters include task complexity , task execution distance and task completion time requirements , the robot ability parameters include the robot processing ability index , current battery level and maximum speed .

[0027] Combine the task complexity parameters and the robot ability parameters to design a state vector, and denote the state vector as , where .

[0028] Set the reward function according to the task completion efficiency and resource consumption. The reward formula is as follows:

[0029] ;

[0030] where represents the task allocation strategy selected by the robot, represents the total energy of the robot, represents the actual time to complete the task, and represent the adjustment weight coefficients, is the energy consumed to complete the task.

[0031] Record the state, action, reward, and next state of each round of task allocation to form an experience pool, and use the Q-learning update formula to optimize the model to minimize the error between the policy and the expected value. The loss function is as follows:

[0032] ;

[0033] where r represents the immediate reward, represents the discount factor, represents the decay of future rewards, represents the model parameters, s represents the current state of the robot during task execution, a represents the action selected by the robot in the current state, is the expected function.

[0034] Evaluate the model performance through the task completion rate, average resource consumption, and task allocation time. Apply the trained model to the real-time task allocation scenario, and dynamically allocate robots and tasks according to the input of the current state .

[0035] Use the reinforcement learning model to analyze the matching situation between task complexity and robot ability in real time, and dynamically allocate tasks to the most suitable robots to improve resource utilization and work efficiency, and ensure the efficient cooperation of each robot.

[0036] As a preferred solution of the industrial robot control method based on deep learning according to the present invention, wherein: the operating state data of the robot are collected in real time, including position, speed, load and energy consumption.

[0037] The data analysis algorithm is adopted to analyze the real-time monitoring data, identify the states deviating from the normal operation mode, and form an anomaly detection mechanism, which is specifically as follows:

[0038] An anomaly detection discriminant function is established , which is used to mark whether the current operating state is abnormal. The function is as follows:

[0039] ;

[0040] Wherein, represents the load of the robot at time t, represents the maximum allowable value of the robot load, represents the speed of the robot at time t, represents the maximum allowable value of the robot speed, represents the energy consumption of the robot at time t, represents the upper limit of the energy consumption per unit time of the robot.

[0041] When , it means that an anomaly is detected.

[0042] When , the state is normal.

[0043] As a preferred solution of the industrial robot control method based on deep learning according to the present invention, wherein: an anomaly warning mechanism is established, which is specifically as follows:

[0044] When is detected, the system triggers a warning signal , and records the anomaly type, anomaly parameters and detection time.

[0045] According to the anomaly type and the degree of parameter overrun, the warning level is set.

[0046] ;

[0047] The and information is fed back to the control system to determine whether it is necessary to reallocate tasks or adjust operating parameters.

[0048] Second aspect, the embodiment of the present invention provides an industrial robot control system based on deep learning, which includes an operation data acquisition module for comprehensively monitoring the operation status and working environment of the robot by installing position, speed, torque, and environment sensors on the industrial robot, regularly collecting real-time data of the robot when performing tasks, and storing the data in a database; a dynamic parameter control module for extracting features from the real-time data, and the extracted features include load, speed, environmental temperature, and workpiece characteristic parameters; establishing a dynamic parameter model according to the features, mapping the real-time data to the control parameter adjustment range, and setting key control parameters including speed gain, acceleration limit, and load response time; a reinforcement learning task allocation module for analyzing the current task attributes and extracting task complexity indicators, combining with the robot ability evaluation model to construct a reinforcement learning decision-making model; designing a reward function through task completion efficiency and resource consumption and using Q-learning to optimize the model to generate an optimal task allocation strategy; an anomaly monitoring and warning module for collecting the operation status data of the robot in real time, marking the abnormal status through an anomaly detection discriminant function, and triggering a warning signal and setting a warning level according to the anomaly type and the degree of parameter overrun.

[0049] Third aspect, the embodiment of the present invention provides a computer device, including a memory and a processor, where the memory stores a computer program, and: when the computer program instructions are executed by the processor, the steps of the industrial robot control method based on deep learning as described in the first aspect of the present invention are implemented.

[0050] Fourth aspect, the embodiment of the present invention provides a computer-readable storage medium, on which a computer program is stored, and: when the computer program instructions are executed by the processor, the steps of the industrial robot control method based on deep learning as described in the first aspect of the present invention are implemented.

[0051] The beneficial effects of the present invention are as follows: The present invention collects the operation data of the robot in real time through sensors to ensure that the subsequent deep learning model can be dynamically adjusted based on accurate data. In the control strategy, by dynamically adjusting the control parameters, the robot can maintain the best operation state under different task conditions. Through the reinforcement learning model, the task complexity and the robot's ability are analyzed and intelligently allocated in real time to optimize the task collaboration efficiency and improve the resource utilization rate. At the same time, an anomaly detection and warning mechanism is constructed, which can quickly identify the operation deviation state. When an anomaly occurs, it feeds back to the control system through a hierarchical warning signal, and reallocates tasks or adjusts parameters to ensure safety; significantly improves the adaptability, collaboration ability, and operation safety of industrial robots in complex scenarios, and realizes efficient and intelligent operation management. Description of the Drawings

[0052] To more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0053] Figure 1 It is a schematic diagram of the steps of an industrial robot control method based on deep learning.

[0054] Figure 2 It is a schematic diagram of the structure of an industrial robot control system based on deep learning. Detailed implementation manners

[0055] To make the above objects, features, and advantages of the present invention more obvious and understandable, the following will make a detailed description of the specific implementation manners of the present invention in conjunction with the drawings of the specification.

[0056] In the following description, many specific details are set forth to fully understand the present invention. However, the present invention can also be implemented in other ways different from those described herein. Those skilled in the art can make similar extensions without departing from the connotation of the present invention. Therefore, the present invention is not limited by the specific embodiments disclosed below.

[0057] Secondly, the so-called "one embodiment" or "embodiment" herein refers to a specific feature, structure, or characteristic that can be included in at least one implementation manner of the present invention. The appearances of "in one embodiment" in different places in this specification do not all refer to the same embodiment, nor are they separate or selectively exclusive embodiments from other embodiments. Embodiment

[0058] Referring to Figures 1 - 2 , which is the first embodiment of the present invention. This embodiment provides an industrial robot control method based on deep learning, including,

[0059] S1: Real-time collect the operation data of the robot during the working process through sensors, including position, speed, and load information, to provide a basis for subsequent analysis.

[0060] Install a variety of sensors on the industrial robot, including position sensors, speed sensors, torque sensors, and environmental sensors, to comprehensively monitor the operation status and working environment of the robot; regularly collect the real-time data of the robot when performing tasks, and the real-time data includes position data, speed data, torque data, and environmental data.

[0061] Store the collected data in a database and perform preprocessing, cleaning, and standardizing the data for subsequent analysis.

[0062] S2: Dynamically adjust the control parameters according to the real-time data changes to optimize the operation performance of the robot and ensure effective operation in different workpieces or environments.

[0063] Extract features from the real-time data, where the features include the current load , the current speed , the current environmental temperature and the workpiece characteristic parameters , and the workpiece characteristic parameters include the material parameters and the shape parameters.

[0064] According to the extracted features, establish a dynamic parameter model, map the real-time data to the adjustment range of the control parameters, and define key control parameters including speed gain, acceleration limit, and load response time, etc., to cope with different workpieces and environmental conditions, as follows:

[0065] Based on the features, establish a mapping relationship, convert the extracted features into the adjustment range of the control parameters, and set the control parameters, including speed gain , acceleration limit and load response time , and the formulas for the speed gain , acceleration limit and load response time are respectively:

[0066] ;

[0067] ;

[0068] ;

[0069] where, represents the reference speed gain, represents the coefficient related to the workpiece characteristics, adjusted experimentally, represents the coefficient related to the environmental temperature, adjusted experimentally, represents the reference value of the workpiece characteristics (e.g., the standard hardness of the workpiece), represents the reference value of the environmental temperature (e.g., the set temperature at room temperature), represents the attenuation factor of the control speed, determining the sensitivity of the speed gain to the current speed, represents the reference acceleration limit, represents the load influence coefficient, represents the environmental temperature influence coefficient, represents the reference value of the load upper limit, represents the speed influence coefficient, represents the reference value of the speed upper limit, represents the reference response time, represents the load impact coefficient, represents the environmental temperature impact coefficient, represents the load reference value.

[0070] Apply the dynamic parameter model to the control strategy as follows:

[0071] The control parameters generated using the dynamic parameter model are dynamically adjusted during the operation of the robot. According to the real-time monitoring data, the speed gain and the acceleration limit are adjusted. The adjustment formula is as follows:

[0072] ;

[0073] ;

[0074] where, represents the safety threshold of the preset load, and represent the preset adjustment coefficients;

[0075] By adjusting the control parameters (such as speed gain and acceleration limit) in real time, ensure that the robot maintains the best operating state under different workpieces or environments, thereby optimizing the execution efficiency, as follows:

[0076] ;

[0077] where, is the control output, is the error, , and are the proportional, integral and differential gains respectively.

[0078] Introduce a performance feedback mechanism. By evaluating the actual performance of the robot under different conditions, dynamically update the feature extraction and parameter model to continuously improve the control strategy and achieve the self-learning ability based on historical data.

[0079] S3: Based on the reinforcement learning model, analyze the complexity of the current task and the capabilities of the robots in real time, and intelligently allocate the collaborative tasks among multiple robots to improve the overall work efficiency.

[0080] Specifically, by analyzing the attributes of the current task, extract the task complexity metrics, such as operation time, precision requirements, and workpiece characteristics, to provide the basic data for subsequent task allocation.

[0081] Furthermore, the working status and performance of each robot are monitored in real time, including the current load, energy consumption, and remaining task capabilities, to form a robot ability evaluation model, ensuring that task allocation takes into account the actual operation capabilities of each robot.

[0082] Based on the evaluation results of task complexity and robot capabilities, a reinforcement learning decision model is constructed. The model is trained with historical data to enable intelligent decision-making in various task scenarios and optimize the task allocation strategy. The specific process is as follows:

[0083] Collect and label the characteristics and historical allocation data of each task, including task complexity parameters and robot ability parameters. The task complexity parameters include task complexity , task execution distance and task completion time requirements , and the robot ability parameters include the robot processing ability index , current battery level and maximum speed .

[0084] Combine the task complexity parameters and robot ability parameters to design a state vector, denoted as , where .

[0085] Set a reward function according to the task completion efficiency and resource consumption. The reward formula is as follows:

[0086] ;

[0087] where represents the task allocation strategy selected by the robot, that is, task is assigned to robot , represents the energy consumed to complete the task, represents the total energy of the robot, represents the actual time to complete the task, and represent the adjustment weight coefficients, respectively controlling the importance of time and energy consumption.

[0088] Furthermore, record the state, action, reward, and next state of each round of task allocation to form an experience pool, and use the Q-learning update formula to optimize the model to minimize the error between the policy and the expected value. The loss function is as follows:

[0089] ;

[0090] where r represents the immediate reward, represents the discount factor, Represents the decay of future rewards and the current Q-value estimate. Represents the model parameters, s represents the current state of the robot during task execution, and a represents the action selected by the robot in the current state. Expected function.

[0091] Evaluate the model performance through the task completion rate, average resource consumption, and task allocation time. Apply the trained model to the real-time task allocation scenario, and dynamically allocate robots and tasks according to the input of the current state to achieve optimized decision-making.

[0092] Utilize the reinforcement learning model to analyze the matching situation between task complexity and robot capabilities in real time, and dynamically allocate tasks to the most suitable robots to improve resource utilization and work efficiency, ensuring efficient cooperation among robots.

[0093] S4: Continuously monitor the running state of the robot and issue early warnings in a timely manner when abnormalities are detected to ensure operation safety.

[0094] First, collect the running state data of the robot in real time, including position, speed, load, and energy consumption, etc., to ensure comprehensive monitoring of the robot's running process.

[0095] Adopt data analysis algorithms to analyze the real-time monitoring data, identify the states that deviate from the normal running mode, such as overloaded load, abnormal speed, or abnormal energy consumption, and form an anomaly detection mechanism as follows:

[0096] Establish an anomaly detection discrimination function , which is used to mark whether the current running state is abnormal. The function is as follows:

[0097] ;

[0098] Among them, represents the load of the robot at time t, represents the maximum allowable value of the robot's load, represents the speed of the robot at time t, represents the maximum allowable value of the robot's speed, represents the energy consumption of the robot at time t, represents the upper limit of the robot's energy consumption per unit time.

[0099] When it means an anomaly is detected.

[0100] When the state is normal.

[0101] Furthermore, establish an anomaly early warning mechanism as follows:

[0102] When it is detected that the system triggers a warning signal and records the type of anomaly, anomaly parameters, and detection time. For example, for different types of anomalies, different warning levels can be set such as "load exceeded" and "abnormal energy consumption" can be set to a high priority level.

[0103] Based on the type of anomaly and the degree of parameter exceedance, set the warning level so as to take different levels of response measures:

[0104] ;

[0105] Feed and information back to the control system to facilitate the management system to judge whether it is necessary to reallocate tasks or adjust operating parameters.

[0106] Furthermore, this embodiment also provides an industrial robot control system based on deep learning, including a running data acquisition module, which is used to comprehensively monitor the running state and working environment of the robot by installing various sensors such as position, speed, torque, and environmental sensors on the industrial robot; this module will regularly collect real-time data (position, speed, torque, and environmental data) when the robot is performing tasks, and store the data in a database, and complete data cleaning and standardization processing; a dynamic parameter control module, which is used to extract features from the real-time data, and the extracted features include load, speed, environmental temperature, and workpiece characteristic parameters, etc.; establish a dynamic parameter model based on the features, map the real-time data to the control parameter adjustment range, and set key control parameters including speed gain, acceleration limit, and load response time, etc., to ensure that the robot reaches the optimal operation state under different workpieces or environmental conditions; a reinforcement learning task allocation module, which is used to analyze the current task attributes and extract task complexity indicators (such as operation time, accuracy requirements, workpiece characteristics), combined with the robot ability evaluation model (including load, energy consumption, task ability, etc.), to construct a reinforcement learning decision-making model. Design a reward function through task completion efficiency and resource consumption and use Q-learning to optimize the model to generate an optimal task allocation strategy; an anomaly monitoring and warning module, which is used to collect real-time running state data of the robot (including position, speed, load, and energy consumption) in real time, and use data analysis algorithms to detect states that deviate from the normal operation mode such as load exceeded, speed anomaly, and energy consumption anomaly. Mark the anomaly state through the anomaly detection discriminant function, and trigger a warning signal and set the warning level according to the type of anomaly and the degree of parameter exceedance, and feedback the warning to the control system to adjust the task allocation or control parameters.

[0107] This embodiment also provides a computer device applicable to the case of an industrial robot control method based on deep learning, including a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement the industrial robot control method based on deep learning proposed in the above embodiment.

[0108] The computer device may be a terminal, and the computer device includes a processor, a memory, a communication interface, a display screen, and an input device connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be implemented through WIFI, a carrier network, NFC (Near Field Communication), or other technologies. The display screen of the computer device may be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device may be a touch layer covering the display screen, or a button, a trackball, or a touchpad provided on the outer shell of the computer device, or an external keyboard, touchpad, or mouse, etc.

[0109] This embodiment also provides a storage medium on which a computer program is stored, and when the program is executed by a processor, it implements the industrial robot control method based on deep learning proposed in the above embodiment.

[0110] In summary, the present invention collects robot operation data in real time through sensors to ensure that the subsequent deep learning model can be dynamically adjusted based on accurate data. In the control strategy, by dynamically adjusting control parameters, the robot can maintain the best operating state under different task conditions. Through the reinforcement learning model, the task complexity and the robot's capabilities are analyzed and intelligently allocated in real time to optimize the task collaboration efficiency and improve the resource utilization rate. At the same time, an anomaly detection and warning mechanism is constructed, which can quickly identify the operating deviation state. When an anomaly occurs, a hierarchical warning signal is fed back to the control system, and task reallocation or parameter adjustment is performed to ensure safety; it significantly improves the adaptability, collaboration ability, and operating safety of industrial robots in complex scenarios, and realizes efficient and intelligent operation management. Embodiment

[0111] In the second embodiment, an industrial robot control method based on deep learning is provided. In order to verify the beneficial effects of the present invention, a scientific demonstration is carried out through a simulation experiment.

[0112] The experiment was conducted in a standard industrial environment where multiple industrial robots performed similar assembly tasks. During the experiment, real-time data was collected from each robot at regular intervals and stored in a database. The data collection frequency was set at once per second to ensure that subtle changes in the robot's working state could be captured. After the data was stored, data preprocessing, including cleaning and standardization, was carried out to ensure the accuracy of subsequent analysis.

[0113] Next, according to the changes in the collected real-time data, the control parameters were dynamically adjusted to optimize the robot's operation performance. In the specific implementation process, features such as the current load, speed, environmental temperature, and the material and shape of the workpiece were extracted. Then, based on the extracted features, a dynamic parameter model was established to map the real-time data to the adjustment range of the control parameters. According to these features, control parameters, including speed gain, acceleration limit, and load response time, were set. Relevant coefficients and reference values were defined in the formula to cope with different workpieces and environmental conditions.

[0114] In practical applications, the control parameters generated by the dynamic parameter model were applied to the robot control strategy in real time. When abnormal situations were detected, such as overloaded or abnormal speed, the system would trigger a warning signal and record the abnormal parameters and detection time. By comparing the working states and performances of different robots, the task allocation was dynamically adjusted to ensure that each robot worked in the best state. To verify the effectiveness of this method, the experimental group was compared with the control group, and the differences in task completion rate, energy consumption, and execution time of the robots were recorded.

[0115] Table 1 Data table of the operating state and performance of industrial robots

[0116] Test object name Location (m) Speed (m / s) Load (kg) Energy consumption (kWh) Ambient temperature (°C) Task complexity (score) Robot 1 2.5 1.5 10 0.5 22 5 Robot 2 3.0 1.8 12 0.6 23 7 Robot 3 1.8 1.4 8 0.4 21 4 Robot 4 2.7 1.6 15 0.7 24 6 Robot 5 3.1 1.9 11 0.5 23 5 Robot 6 2.0 1.7 9 0.6 20 4 Robot 7 2.3 1.5 13 0.4 22 6

[0117] According to Table 1, the energy consumption performance of the robots under different load conditions can be observed. For example, Robot 4 has the largest load (15 kg), but its energy consumption (0.7 kWh) is also relatively high; while for robots with smaller loads (such as Robot 3, 8 kg), their energy consumption is only 0.4 kWh. This shows that there is a significant relationship between load control and energy consumption.

[0118] The monitoring results of speed and position data show that when the robot has a higher speed (such as Robot 5, 1.9 m / s), its task complexity score (5) does not exceed that of other robots; on the contrary, robots with slower speeds (such as Robot 3, 1.4 m / s) show greater room for improvement in task complexity scores (score 4). This indicates that optimizing speed gain and acceleration limit is crucial for improving the adaptability of robots in complex task environments.

[0119] By comprehensively analyzing various parameters, it is concluded that the dynamic adjustment control strategy adopted by the present invention has significant advantages; compared with the prior art, this method optimizes the operation efficiency and task allocation strategy of the robot through real-time feedback and data analysis; specifically, in the scenario of multiple robots collaborating, the implemented self-learning ability enables the robot to continuously adjust its own parameters according to historical data, improving the utilization rate of resources and the overall work efficiency.

[0120] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical solutions of the present invention, and they should all be covered within the scope of the claims of the present invention.

Claims

1. An industrial robot control method based on deep learning, characterized in that: Including: S1: Collect the operation data of the robot in real time during the working process through sensors; S2: Dynamically adjust the control parameters according to the changes in real-time data; S3: Based on the reinforcement learning model, analyze the complexity of the current task and the capabilities of the robots in real time, and intelligently allocate the collaborative tasks among multiple robots; S4: Continuously monitor the operation status of the robot and issue a warning when an anomaly is detected; The specific implementation process of S2 includes: Extract features from the real-time data, where the features include the current load , the current speed , the current ambient temperature and the workpiece characteristic parameters , where the workpiece characteristic parameters include material parameters and shape parameters; Establish a dynamic parameter model according to the extracted features, map the real-time data to the adjustment range of the control parameters, and define the key control parameters, where the key control parameters include speed gain, acceleration limit, and load response time; Establish a mapping relationship based on features, convert the extracted features into the adjustment range of control parameters, and set control parameters, including speed gain , acceleration limit and load response time , the speed gain , acceleration limit and load response time The formulas for are as follows: ； ； ； Among them, represents the reference speed gain, represents the workpiece characteristic coefficient, represents the ambient temperature coefficient, represents the reference value of workpiece characteristics, represents the reference value of ambient temperature, represents the attenuation factor for controlling speed, represents the reference acceleration limit, represents the load influence coefficient, represents the ambient temperature influence coefficient, represents the reference value of the load upper limit, represents the speed influence coefficient, represents the reference value of the speed upper limit, represents the reference response time, represents the load influence coefficient, represents the ambient temperature influence coefficient, represents the reference value of the load; S2 also includes the following steps: Apply the dynamic parameter model to the control strategy, specifically as follows: The control parameters generated using the dynamic parameter model are dynamically adjusted during the operation of the robot. According to the real-time monitoring data, the speed gain is adjusted and the acceleration limit , and the adjustment formula is as follows: ； ； Among them, represents the safety threshold of the preset load, and represents the preset adjustment coefficient; Introduce a performance feedback mechanism, evaluate the actual performance of the robot under different conditions, dynamically update the feature extraction and parameter model, and achieve the self-learning ability based on historical data.

2. The industrial robot control method based on deep learning according to claim 1, characterized in that: The specific implementation process of step S1 includes: Install a variety of sensors on the industrial robot, including position sensors, speed sensors, torque sensors, and environmental sensors; Regularly collect the real-time data of the robot when performing tasks, where the real-time data includes position data, speed data, torque data, and environmental data; Store the collected data in the database and perform preprocessing to clean and standardize the data.

3. The industrial robot control method based on deep learning according to claim 2, wherein: The specific implementation process of S3 includes: Extract the task complexity index by analyzing the attributes of the current task; Real-time monitor the working status and performance of each robot, including the current load, energy consumption, and remaining task capabilities, and form a robot ability evaluation model; Based on the evaluation results of task complexity and robot capabilities, construct a reinforcement learning decision model, train the model through historical data, and optimize the task allocation strategy. The specific process is as follows: Collect and label the characteristics and historical allocation data of each task, including task complexity parameters and robot ability parameters. The task complexity parameters include task complexity , task execution distance and task completion time requirement . The robot ability parameters include robot processing ability index , current battery level and maximum speed ; Combine the task complexity parameter and the robot ability parameter to design a state vector, and denote the state vector as , where ; Set the reward function according to the task completion efficiency and resource consumption. The reward formula is as follows: ； Among them, represents the task assignment strategy selected by the robot, represents the total energy of the robot, represents the time actually taken to complete the task, and represents the adjustment weight coefficient, is the energy consumed to complete the task; Record the state, action, reward, and next state of each round of task allocation to form an experience pool, and use the Q-learning update formula to optimize the model to minimize the error between the policy and the expected value. The loss function is as follows: ； where r represents the immediate reward, represents the discount factor, indicating the decay of future rewards, represents the model parameters, s represents the current state of the robot during task execution, and a represents the action selected by the robot in the current state, expected function; Evaluate the model performance through the task completion rate, average resource consumption, and task allocation time, apply the trained model to the real-time task allocation scenario, and dynamically allocate robots and tasks according to the input of the current state ; Use the reinforcement learning model to analyze the matching situation between task complexity and robot capabilities in real time, dynamically allocate tasks to the most suitable robot to improve resource utilization and work efficiency, and ensure the efficient cooperation of each robot.

4. The industrial robot control method based on deep learning according to claim 3, wherein: The specific implementation process of S4 includes: Collect the operation status data of the robot in real time, including position, speed, load, and energy consumption; Adopt a data analysis algorithm to analyze the real-time monitoring data, identify the status deviating from the normal operation mode, and form an anomaly detection mechanism, specifically as follows: Establish an anomaly detection discrimination function , which is used to mark whether the current running state is abnormal. The function is as follows: ； Among them, represents the load of the robot at time t, represents the maximum allowable value of the robot's load, represents the speed of the robot at time t, represents the maximum allowable value of the robot's speed, represents the energy consumption of the robot at time t, represents the upper limit of the robot's energy consumption per unit time; When it indicates that an abnormality has been detected; When the status is normal.

5. The industrial robot control method based on deep learning according to claim 4, wherein: Step S4 also includes the following steps: Establish an anomaly warning mechanism, specifically as follows: When detected the system triggers a warning signal and records the type of anomaly, anomaly parameters, and detection time; Set the warning level according to the anomaly type and the degree of parameter overrun: ； Feed back the information to the control system to determine whether it is necessary to reallocate tasks or adjust the operating parameters.

6. An industrial robot control system based on deep learning, based on the industrial robot control method based on deep learning according to any one of claims 1 to 5, characterized in that: Also included, An operation data acquisition module, which is used to comprehensively monitor the operation status and working environment of the robot by installing sensors on the industrial robot, regularly collect the real-time data of the robot when performing tasks, and store the data in the database; The dynamic parameter control module is used to extract features from real-time data, extract characteristic parameters, establish a dynamic parameter model based on the features, map the real-time data to the control parameter adjustment range, and set key control parameters; The reinforcement learning task allocation module is used to analyze the current task attributes and extract task complexity indicators, combine with the robot ability evaluation model, construct a reinforcement learning decision model, design a reward function through task completion efficiency and resource consumption, and finally generate an optimal task allocation strategy; The anomaly monitoring and warning module is used to collect the operation status data of the robot in real time, mark the abnormal status through the anomaly detection discriminant function, and trigger a warning signal and set a warning level according to the anomaly type and the degree of parameter overrun.

7. A computer device, comprising a memory and a processor, the memory storing a computer program, characterized in that: When the processor executes the computer program, it implements the steps of the industrial robot control method based on deep learning according to any one of claims 1 to 5.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by the processor, it implements the steps of the industrial robot control method based on deep learning according to any one of claims 1 to 5.

Images