Energy consumption optimization control method for green intelligent manufacturing body-equipped robot

Abstract

The invention relates to the technical field of robots, in particular to an energy consumption optimization control method for a green intelligent manufacturing body-equipped robot, and the method comprises the steps: collecting the working condition state data of the body-equipped robot in the process of executing a manufacturing task, and carrying out the preprocessing of the working condition state data, so as to obtain a state feature; performing mapping alignment in combination with process behaviors and state features in the manufacturing task to obtain a process energy consumption feature set corresponding to the process behaviors; obtaining a constraint condition of a manufacturing task, and training the initial model based on the process energy consumption feature set and the constraint condition to obtain an energy consumption optimization model; based on the periodic energy consumption standard deviation of the manufacturing task, whether the energy-saving control stability of the robot with the body meets the requirement or not is determined; the energy consumption constraint disturbance tolerance coefficient of the energy consumption optimization model is determined according to the motor operation current fluctuation rate of the robot with the body; and determining a residual feedback gain of energy consumption prediction based on the success rate of the manufacturing task. The energy-saving control stability of the robot with the body is improved.

Description

A method for optimizing energy consumption control of green intelligent manufacturing embodied robots Technical Field

[0001] This invention relates to the field of robotics, and in particular to a method for optimizing and controlling the energy consumption of a green intelligent manufacturing embodied robot. Background Technology

[0002] Against the backdrop of the accelerated development of green manufacturing and smart factories, energy consumption optimization control methods for embodied robots in green intelligent manufacturing have become an important technological direction for improving the energy utilization efficiency of manufacturing systems, reducing energy consumption per unit output, and achieving sustainable production. Because embodied robots involve multi-joint collaborative motion and multi-task parallel scheduling during manufacturing tasks, their energy consumption performance depends not only on a single motor or control parameter, but also on the combined influence of multiple factors such as task decomposition strategies, motion control algorithms, energy recovery mechanisms, and production cycle coordination, directly affecting production costs and carbon emission levels. Existing robot control systems mostly focus on optimizing motion accuracy and cycle efficiency. While they possess basic energy consumption monitoring and limit control functions, they still have shortcomings in cross-task energy consumption coupling modeling, multi-objective constraint collaborative optimization, and adaptive adjustment capabilities to operating condition disturbances. This makes it difficult to achieve dynamic optimal energy consumption control while ensuring processing quality and production cycle time. Therefore, there is an urgent need for a green intelligent manufacturing embodied robot energy consumption optimization control method that can integrate energy consumption prediction models, constraint disturbance suppression mechanisms and multi-objective collaborative optimization algorithms to achieve refined energy consumption scheduling during task execution and dynamic balance between energy-saving goals and production efficiency.

[0003] Chinese Patent Publication No. CN110936382A discloses a data-driven method for optimizing the energy consumption of industrial robots, comprising: Step 1, determining the parameters affecting the energy consumption of industrial robots and constructing a mathematical relationship equation between the energy consumption of industrial robots and the influencing parameters; Step 2, establishing a neural network model to describe the relationship between the energy consumption of industrial robots and the influencing parameters; training the inter-layer weights and thresholds of each layer of the established neural network model to obtain a trained neural network model; Step 3, using the energy consumption of industrial robots as the optimization objective and using non-optimized parameters as fixed inputs to the trained neural network model to optimize the motion parameters of industrial robots. It is evident that the data-driven method for optimizing the energy consumption of industrial robots suffers from problems such as insufficient coverage of training data, inadequate reflection of diverse operating conditions, and accumulated energy consumption prediction deviations leading to insufficient stability in the energy-saving control of the robot. Summary of the Invention

[0004] To address this issue, the present invention provides a green intelligent manufacturing embodied robot energy consumption optimization control method to overcome the problem in the prior art where the limited coverage of training data and the lack of full reflection of the diversity of working conditions lead to the accumulation of energy consumption prediction deviations, resulting in insufficient stability of energy-saving control of embodied robots.

[0005] To achieve the above objectives, this invention provides a green intelligent manufacturing embodied robot energy consumption optimization and control method, comprising: collecting working condition state data of the embodied robot during the execution of a manufacturing task; preprocessing the working condition state data to obtain state features; mapping and aligning the process behaviors in the manufacturing task with the state features to obtain a process energy consumption feature set corresponding to the process behaviors; obtaining the constraints of the manufacturing task; training an initial model based on the process energy consumption feature set and the constraints to obtain an energy consumption optimization model; generating an energy consumption optimization strategy based on the energy consumption optimization model to optimize the energy consumption of the embodied robot; and obtaining the manufacturing task... The standard deviation of the cycle energy consumption of the manufacturing task is used to determine whether the energy-saving control stability of the embodied robot meets the requirements. If the energy-saving control stability of the embodied robot does not meet the requirements, the fluctuation rate of the motor operating current of the embodied robot is obtained to determine whether the robustness of the control strategy meets the requirements. If the robustness of the control strategy does not meet the requirements, the fluctuation rate of the motor operating current of the embodied robot is used to determine whether to reduce the energy consumption constraint disturbance tolerance coefficient of the energy consumption optimization model. If it is not necessary to reduce the energy consumption constraint disturbance tolerance coefficient of the energy consumption optimization model, the success rate of the manufacturing task is used to determine whether to increase the residual feedback gain of energy consumption prediction.

[0006] Further, determining whether the energy-saving control stability of the embodied robot meets the requirements based on the standard deviation of the cycle energy consumption of the manufacturing task includes: comparing the standard deviation of the cycle energy consumption of the manufacturing task with a preset standard deviation of energy consumption; if the standard deviation of the cycle energy consumption of the manufacturing task is less than or equal to the preset standard deviation of energy consumption, then the energy-saving control stability of the embodied robot meets the requirements; if the standard deviation of the cycle energy consumption of the manufacturing task is greater than the preset standard deviation of energy consumption, then the energy-saving control stability of the embodied robot does not meet the requirements.

[0007] Furthermore, given that the energy-saving control stability of the avatar robot does not meet the requirements, the robustness of the control strategy is determined based on the fluctuation rate of the motor operating current of the avatar robot.

[0008] Furthermore, determining whether the robustness of the control strategy meets the requirements based on the fluctuation rate of the motor operating current of the embodied robot includes: comparing the fluctuation rate of the motor operating current of the embodied robot with a preset first fluctuation rate; if the fluctuation rate of the motor operating current of the embodied robot is less than or equal to the preset first fluctuation rate, then the robustness of the control strategy meets the requirements; if the fluctuation rate of the motor operating current of the embodied robot is greater than the preset first fluctuation rate, then the robustness of the control strategy does not meet the requirements.

[0009] Further, determining whether to reduce the energy consumption constraint disturbance tolerance coefficient of the energy consumption optimization model includes: comparing the fluctuation rate of the motor operating current of the android with the preset first fluctuation rate and the preset second fluctuation rate respectively; if the fluctuation rate of the motor operating current of the android is greater than the preset first fluctuation rate and less than or equal to the preset second fluctuation rate, reducing the energy consumption constraint disturbance tolerance coefficient of the energy consumption optimization model; if the fluctuation rate of the motor operating current of the android is greater than the preset second fluctuation rate, determining that it is not necessary to reduce the energy consumption constraint disturbance tolerance coefficient of the energy consumption optimization model.

[0010] Furthermore, the reduction in the energy consumption constraint disturbance tolerance coefficient of the energy consumption optimization model is determined by the difference between the motor operating current fluctuation rate of the embodied robot and the preset first fluctuation rate.

[0011] Furthermore, based on the condition that the fluctuation rate of the motor operating current of the embodied robot is greater than the preset second fluctuation rate, it is initially determined that the accuracy of the energy consumption optimization model in predicting energy consumption does not meet the requirements, and the accuracy of the energy consumption prediction of the energy consumption optimization model is determined based on the success rate of the manufacturing task.

[0012] Furthermore, the accuracy of the energy consumption prediction of the energy consumption optimization model is determined based on the success rate of the manufacturing task, including: comparing the success rate of the manufacturing task with a preset success rate; if the success rate of the manufacturing task is greater than the preset success rate, the accuracy of the energy consumption prediction of the energy consumption optimization model is determined to meet the requirements; if the success rate of the manufacturing task is less than or equal to the preset success rate, the accuracy of the energy consumption prediction of the energy consumption optimization model is determined to not meet the requirements, and the residual feedback gain of the energy consumption prediction is increased.

[0013] Furthermore, the success rate of the manufacturing task is the ratio of the number of times the manufacturing task is successfully completed during the execution of the manufacturing task to the total number of manufacturing operations.

[0014] Furthermore, the increase in the residual feedback gain of the energy consumption prediction is determined by the difference between the preset success rate and the success rate of the manufacturing task.

[0015] Compared with the prior art, the beneficial effects of the present invention are as follows: The method of the present invention determines whether the energy-saving control stability of the embodied robot meets the requirements based on the standard deviation of the cycle energy consumption of the manufacturing task. Since the energy-saving control strategy is affected by various factors, the difference in the complexity of different tasks leads to large fluctuations in cycle energy consumption. By determining the energy-saving control stability of the embodied robot, the response consistency of the energy-saving strategy under different task conditions can be quantified, and task segments with large energy consumption fluctuations can be identified, providing a basis for optimizing control parameters and improving the overall energy-saving effect. The energy consumption constraint disturbance tolerance coefficient of the energy consumption task coupling model is adjusted based on the fluctuation rate of the motor operating current of the embodied robot. During task execution, the motor load is affected by the combined influence of multiple factors such as process complexity and environmental disturbances, leading to changes in operating... Significant fluctuations in current increase the frequency of control strategy adjustments, leading to energy consumption peaks and mechanical shocks. Reducing the energy consumption constraint disturbance tolerance coefficient can decrease the model's sensitivity to instantaneous fluctuations, enabling the coupled model to maintain a stable response when executing multiple energy consumption optimization objectives, reducing energy consumption fluctuations and protecting the execution components. The residual feedback gain of energy consumption prediction is adjusted based on the success rate of the manufacturing task. Because the energy consumption prediction model fails to fully capture the complex energy consumption characteristics during task execution, prediction errors accumulate and affect subsequent control decisions. Increasing the residual feedback gain enhances the model's response to prediction errors, allowing the energy consumption optimization strategy to quickly correct deviations, improving prediction accuracy, and thus ensuring the reliability of the energy consumption optimization strategy execution, thereby improving the energy-saving control stability of the embodied robot.

[0016] Furthermore, this invention determines whether the energy-saving control stability of the embodied robot meets the requirements by setting a preset energy consumption standard deviation. Since the energy-saving control strategy is affected by various factors, the difference in the complexity of different tasks leads to large fluctuations in cycle energy consumption. By determining the energy-saving control stability of the embodied robot, the response consistency of the energy-saving strategy under different task conditions can be quantified, and task links with large energy consumption fluctuations can be identified. This provides a basis for optimizing control parameters and improving the overall energy-saving effect, further improving the energy-saving control stability of the embodied robot.

[0017] Furthermore, this invention adjusts the energy consumption constraint disturbance tolerance coefficient of the energy consumption task coupling model by setting a preset first volatility and a preset second volatility. During task execution, the motor load is affected by multiple factors such as process complexity and environmental disturbance, resulting in large fluctuations in the operating current. Frequent current fluctuations increase the frequency of control strategy adjustments, causing energy consumption peaks and mechanical shocks. By reducing the energy consumption constraint disturbance tolerance coefficient, the sensitivity of the model to instantaneous fluctuations can be reduced, enabling the coupling model to maintain a stable response when executing multiple energy consumption optimization objectives, reducing energy consumption fluctuations and protecting the execution components, thereby further improving the energy-saving control stability of the embodied robot.

[0018] Furthermore, this invention adjusts the residual feedback gain of energy consumption prediction by setting a preset success rate. Since the energy consumption prediction model fails to fully capture the complex energy consumption characteristics during task execution, prediction errors accumulate and affect subsequent control decisions. By increasing the residual feedback gain, the model's response to prediction errors can be enhanced, enabling the energy consumption optimization strategy to quickly correct deviations, improve prediction accuracy, and thus ensure the reliability of the energy consumption optimization strategy execution, further improving the energy-saving control stability of the embodied robot. Attached Figure Description

[0019] Figure 1 is an overall flowchart of the energy consumption optimization control method for green intelligent manufacturing embodied robots according to an embodiment of the present invention; Figure 2 is a logical flowchart of the process of determining whether the energy-saving control stability of the embodied robot meets the requirements in the energy consumption optimization control method for green intelligent manufacturing embodied robots according to an embodiment of the present invention; Figure 3 is a logical flowchart of the process of determining whether to reduce the energy consumption constraint disturbance tolerance coefficient of the energy consumption optimization model in the energy consumption optimization control method for green intelligent manufacturing embodied robots according to an embodiment of the present invention; Figure 4 is a logical flowchart of the process of determining whether to increase the residual feedback gain of energy consumption prediction in the energy consumption optimization control method for green intelligent manufacturing embodied robots according to an embodiment of the present invention. Detailed Implementation

[0020] To make the objectives and advantages of the present invention clearer, the present invention will be further described below with reference to embodiments; it should be understood that the specific embodiments described herein are merely for explaining the present invention and are not intended to limit the present invention.

[0021] Preferred embodiments of the present invention will now be described with reference to the accompanying drawings. Those skilled in the art should understand that these embodiments are merely illustrative of the technical principles of the present invention and are not intended to limit the scope of protection of the present invention.

[0022] Please refer to Figure 1, which is an overall flowchart of the energy consumption optimization control method for green intelligent manufacturing embodied robots according to an embodiment of the present invention.

[0023] This invention discloses a green intelligent manufacturing embodied robot energy consumption optimization and control method, comprising: Step S1, collecting working condition state data of the embodied robot during the execution of a manufacturing task, preprocessing the working condition state data to obtain state features, and mapping and aligning the process behaviors in the manufacturing task with the state features to obtain a process energy consumption feature set corresponding to the process behaviors; Step S2, obtaining the constraints of the manufacturing task, training an initial model based on the process energy consumption feature set and the constraints to obtain an energy consumption optimization model, and generating an energy consumption optimization strategy based on the energy consumption optimization model to optimize the energy consumption of the embodied robot; Step S3, obtaining the cycle energy of the manufacturing task. Step S4: If the energy-saving control stability of the embodied robot does not meet the requirements, the motor operating current fluctuation rate of the embodied robot is obtained to determine whether the robustness of the control strategy meets the requirements. Step S5: If the robustness of the control strategy does not meet the requirements, the energy consumption constraint disturbance tolerance coefficient of the energy consumption optimization model is determined based on the motor operating current fluctuation rate of the embodied robot. Step S6: If it is not necessary to reduce the energy consumption constraint disturbance tolerance coefficient of the energy consumption optimization model, the residual feedback gain of energy consumption prediction is determined based on the success rate of the manufacturing task.

[0024] Specifically, the manufacturing tasks include the production of hair care products, sun protection products, and facial cleansing products.

[0025] Specifically, the operating status data includes motor operation data, energy consumption data, and task progress data.

[0026] Specifically, preprocessing includes cleaning, denoising, deduplication, normalization, and feature extraction.

[0027] Specifically, the state characteristics include current fluctuation rate, energy consumption trend characteristics, and mission success rate.

[0028] Specifically, the process activities include weighing, filling, and packaging.

[0029] Specifically, the process of mapping and aligning state features according to process behaviors in manufacturing tasks to divide the process energy consumption feature set corresponding to process behaviors involves mapping and aligning state features with corresponding process behaviors according to timestamps. On this basis, the state features in each process interval are statistically or filtered to obtain a feature set that reflects the energy consumption performance and action characteristics of the corresponding process, i.e., the process energy consumption feature set.

[0030] Specifically, the process energy consumption characteristic set includes the weighing process characteristic set, the filling process characteristic set, and the packaging process characteristic set.

[0031] Specifically, the constraints include upper limit constraints on energy consumption per process, constraints on task completion time, and upper limit constraints on motor current.

[0032] Specifically, the upper limit constraint on energy consumption of a single process is the maximum total energy consumption allowed for a single process.

[0033] Specifically, the task completion time constraint is the maximum processing time allowed for a single operation.

[0034] Specifically, the upper limit constraint on motor current is the maximum allowable operating current of the robot's motor during the execution of manufacturing tasks.

[0035] Specifically, the process of training an initial model to obtain an energy consumption optimization model based on the process energy consumption feature set and constraints is as follows: the process energy consumption feature set is used as the input of the initial model, and the constraints are used as the boundary or loss function constraints for model training. The initial model is iteratively trained and the parameters are optimized. By continuously adjusting the model weights, it can predict and optimize the energy consumption of each process and the overall task under the premise of satisfying the constraints, so as to achieve the optimal or near-optimal control strategy for energy consumption, and thus output the energy consumption optimization model.

[0036] Specifically, the initial model is a model framework with the ability to learn and predict nonlinear mappings between tasks and energy consumption.

[0037] Specifically, the energy consumption optimization model can be a random forest regression model, a reinforcement learning model, or a gray box model, with the preferred embodiment being the gray box model.

[0038] Specifically, energy consumption optimization strategies include power allocation strategies, task scheduling strategies, and action adjustment strategies.

[0039] Specifically, the process of generating an energy consumption optimization strategy based on an energy consumption optimization model to optimize the energy consumption of the embodied robot involves simulating or evaluating possible action schemes for each process or task batch based on the energy consumption optimization model, calculating the energy consumption index and stability index of each scheme under the constraint conditions, selecting or generating the optimal combination of action parameters, process sequence and power allocation scheme based on the evaluation results, forming an energy consumption optimization strategy that can be used to control the embodied robot to perform manufacturing tasks, thereby minimizing the total energy consumption of the embodied robot.

[0040] Specifically, the energy consumption constraint perturbation tolerance coefficient of the energy consumption optimization model is the ratio of the maximum allowable actual energy consumption deviation value to the energy consumption constraint value during the energy consumption optimization process.

[0041] Specifically, the residual feedback gain of energy consumption prediction is the gain coefficient used to correct the control strategy by feeding back the deviation between actual energy consumption and model-predicted energy consumption.

[0042] In implementation, the method of this invention determines whether the energy-saving control stability of the embodied robot meets the requirements based on the standard deviation of the cycle energy consumption of the manufacturing task. Since the energy-saving control strategy is affected by various factors, the difference in the complexity of different tasks leads to large fluctuations in cycle energy consumption. By determining the energy-saving control stability of the embodied robot, the response consistency of the energy-saving strategy under different task conditions can be quantified, and task segments with large energy consumption fluctuations can be identified. This provides a basis for optimizing control parameters and improving the overall energy-saving effect. The energy consumption constraint disturbance tolerance coefficient of the energy consumption task coupling model is adjusted based on the fluctuation rate of the motor operating current of the embodied robot. During task execution, the motor load is affected by multiple factors such as process complexity and environmental disturbances, resulting in large fluctuations in the operating current. Frequent current fluctuations increase the frequency of control strategy adjustments, leading to energy consumption peaks and mechanical shocks. By reducing the energy consumption constraint disturbance tolerance coefficient, the model's sensitivity to instantaneous fluctuations can be reduced, enabling the coupled model to maintain a stable response when executing multiple energy consumption optimization objectives, reducing energy consumption fluctuations and protecting the execution components. The residual feedback gain of energy consumption prediction is adjusted according to the success rate of the manufacturing task. Since the energy consumption prediction model fails to fully capture the complex energy consumption characteristics during task execution, prediction errors accumulate and affect subsequent control decisions. By increasing the residual feedback gain, the model's response to prediction errors can be enhanced, enabling the energy consumption optimization strategy to quickly correct deviations, improving prediction accuracy, and thus ensuring the reliability of the energy consumption optimization strategy execution, thereby improving the energy-saving control stability of the embodied robot.

[0043] Please refer to Figure 2, which is a flowchart illustrating the process of determining whether the energy-saving control stability of the embodied robot meets the requirements in the green intelligent manufacturing embodied robot energy consumption optimization control method of this embodiment of the invention.

[0044] Specifically, determining whether the energy-saving control stability of the embodied robot meets the requirements based on the standard deviation of the cycle energy consumption of the manufacturing task includes: comparing the standard deviation of the cycle energy consumption of the manufacturing task with a preset standard deviation of energy consumption; if the standard deviation of the cycle energy consumption of the manufacturing task is less than or equal to the preset standard deviation of energy consumption, then the energy-saving control stability of the embodied robot meets the requirements; if the standard deviation of the cycle energy consumption of the manufacturing task is greater than the preset standard deviation of energy consumption, then the energy-saving control stability of the embodied robot does not meet the requirements.

[0045] Understandably, in the energy consumption optimization control method for embodied robots, a preset energy consumption standard deviation is used to characterize the energy-saving control stability of the embodied robot. The core logic is to transform energy consumption fluctuations into a quantifiable standard deviation range, and to judge the stability of energy consumption optimization control by comparing the actual periodic energy consumption standard deviation with the preset standard deviation. The preset energy consumption standard deviation can be set according to the actual working conditions. The setting of the preset energy consumption standard deviation aims to ensure the stability and practicality of the energy-saving control of the embodied robot. Optionally, the preset energy consumption standard deviation is determined through a limited number of experiments by evaluating the energy-saving control effect of different energy consumption standard deviations on the embodied robot. The determined preset energy consumption standard deviation should satisfy the condition that it is neither too small nor will it cause excessive interference to the energy-saving control process of the embodied robot. For example, the preset energy consumption standard deviation is generally selected in the range of [4%, 6%].

[0046] Preferably, the preset energy consumption standard deviation is 5% in the preferred embodiment.

[0047] Specifically, the standard deviation of cycle energy consumption for manufacturing tasks for: Where n is the number of execution cycles for the same manufacturing task, and n≥2; E i E represents the energy consumption value for the i-th cycle. 平 This represents the average periodic energy consumption value.

[0048] In practice, this invention uses a preset energy consumption standard deviation to determine whether the energy-saving control stability of the embodied robot meets the requirements. Since the energy-saving control strategy is affected by various factors, the difference in the complexity of different tasks leads to large fluctuations in periodic energy consumption. By determining the energy-saving control stability of the embodied robot, the response consistency of the energy-saving strategy under different task conditions can be quantified, and task links with large energy consumption fluctuations can be identified. This provides a basis for optimizing control parameters and improving the overall energy-saving effect, further improving the energy-saving control stability of the embodied robot.

[0049] Please refer to Figure 3, which is a flowchart illustrating the process of determining whether to reduce the energy consumption constraint disturbance tolerance coefficient of the energy consumption optimization model using the green intelligent manufacturing embodied robot energy consumption optimization control method according to an embodiment of the present invention.

[0050] Specifically, given that the energy-saving control stability of the avatar robot does not meet the requirements, the robustness of the control strategy is determined based on the fluctuation rate of the motor operating current of the avatar robot.

[0051] Specifically, determining whether the robustness of the control strategy meets the requirements based on the fluctuation rate of the motor operating current of the embodied robot includes: comparing the fluctuation rate of the motor operating current of the embodied robot with a preset first fluctuation rate; if the fluctuation rate of the motor operating current of the embodied robot is less than or equal to the preset first fluctuation rate, then the robustness of the control strategy meets the requirements; if the fluctuation rate of the motor operating current of the embodied robot is greater than the preset first fluctuation rate, then the robustness of the control strategy does not meet the requirements.

[0052] Specifically, determining whether to reduce the energy consumption constraint disturbance tolerance coefficient of the energy consumption optimization model includes: comparing the fluctuation rate of the motor operating current of the android with the preset first fluctuation rate and the preset second fluctuation rate respectively; if the fluctuation rate of the motor operating current of the android is greater than the preset first fluctuation rate and less than or equal to the preset second fluctuation rate, reducing the energy consumption constraint disturbance tolerance coefficient of the energy consumption optimization model; if the fluctuation rate of the motor operating current of the android is greater than the preset second fluctuation rate, determining that it is not necessary to reduce the energy consumption constraint disturbance tolerance coefficient of the energy consumption optimization model.

[0053] Understandably, the three intervals defined by the preset first volatility rate and the preset second volatility rate correspond to three different scenarios: The first interval is when the embodied robot's motor operating current volatility is less than or equal to the preset first volatility rate, indicating that the robustness of the control strategy meets the requirements. The second interval is when the embodied robot's motor operating current volatility is greater than the preset first volatility rate and less than or equal to the preset second volatility rate, indicating that during task execution, the motor load is affected by multiple factors such as process complexity and environmental disturbances, leading to significant fluctuations in the operating current. Frequent current fluctuations increase the frequency of control strategy adjustments, causing energy consumption peaks and mechanical shocks. The third interval is when the embodied robot's motor operating current volatility is greater than the preset second volatility rate, indicating that the energy consumption prediction model fails to fully capture the complex energy consumption characteristics during task execution, leading to the accumulation of prediction errors and affecting subsequent control decisions.

[0054] Understandably, in the energy consumption optimization control method for embodied robots, the robustness of the control strategy is characterized by using a preset first volatility rate and a preset second volatility rate. The core logic is to transform the abstract robustness judgment into a quantifiable volatility range judgment by establishing a correlation between the motor operating current volatility and the robustness of the control strategy. The preset first volatility rate serves as the dividing line for stability compliance, and the preset second volatility rate serves as the dividing line for the causes of instability failure, thus providing a basis for subsequent stability improvement adjustments. The preset first volatility rate and the preset second volatility rate can be set according to actual working conditions. The setting of the preset first volatility rate and the preset second volatility rate aims to ensure the stability and practicality of the energy-saving control of the embodied robot. Optionally, the preset first volatility rate and the preset second volatility rate are determined through a limited number of experiments by evaluating the energy-saving control effect of different current volatility rates on the embodied robot. The determined preset first volatility rate and the preset second volatility rate should satisfy the condition that they are neither too small nor cause excessive interference to the energy-saving control process of the embodied robot. For example, the preset first volatility is generally selected in the range of [6%, 8%], and the preset second volatility is generally selected in the range of [11%, 13%].

[0055] Preferably, the first volatility is 7% in a preferred embodiment, and the second volatility is 12% in a preferred embodiment.

[0056] Specifically, the variability rate of the motor operating current of the embodied robot is the ratio of the absolute value of the current change of the motor of the embodied robot during the execution of the manufacturing task to the average current value.

[0057] Specifically, the reduction in the energy consumption constraint disturbance tolerance coefficient of the energy consumption optimization model is determined by the difference between the motor operating current fluctuation rate of the embodied robot and the preset first fluctuation rate.

[0058] Specifically, when the difference between the operating current fluctuation rate of the embodied robot's motor and the preset first fluctuation rate is within 2%, the energy consumption constraint disturbance tolerance coefficient of the energy consumption optimization model is reduced to 0.95 times the original value. When the difference between the operating current fluctuation rate of the embodied robot's motor and the preset first fluctuation rate exceeds 2%, on the basis of reducing it to 0.95 times the original value, for every 1% exceeding 2%, the energy consumption constraint disturbance tolerance coefficient of the energy consumption optimization model is reduced by 0.005. For example, when the difference between the operating current fluctuation rate of the embodied robot's motor and the preset first fluctuation rate is 3%, the current energy consumption constraint disturbance tolerance coefficient of the energy consumption optimization model is 0.05, and the reduced energy consumption constraint disturbance tolerance coefficient of the energy consumption optimization model is 0.05×0.95-0.005×1=0.0425.

[0059] In practice, this invention adjusts the energy consumption constraint disturbance tolerance coefficient of the energy consumption task coupling model by setting a preset first volatility and a preset second volatility. During task execution, the motor load is affected by multiple factors such as process complexity and environmental disturbance, resulting in large fluctuations in the operating current. Frequent current fluctuations increase the frequency of control strategy adjustments, causing energy consumption peaks and mechanical shocks. By reducing the energy consumption constraint disturbance tolerance coefficient, the sensitivity of the model to instantaneous fluctuations can be reduced, enabling the coupling model to maintain a stable response when executing multiple energy consumption optimization objectives, reducing energy consumption fluctuations and protecting the execution components, thereby further improving the energy-saving control stability of the embodied robot.

[0060] Please refer to Figure 4, which is a flowchart of the process of determining whether to increase the residual feedback gain of energy consumption prediction in the energy consumption optimization control method of green intelligent manufacturing embodied robot according to an embodiment of the present invention.

[0061] Specifically, based on the condition that the fluctuation rate of the motor operating current of the embodied robot is greater than the preset second fluctuation rate, it is initially determined that the accuracy of the energy consumption optimization model in predicting energy consumption does not meet the requirements, and the accuracy of the energy consumption prediction of the energy consumption optimization model is determined based on the success rate of the manufacturing task.

[0062] Specifically, the accuracy of the energy consumption prediction of the energy consumption optimization model is determined based on the success rate of the manufacturing task, including: comparing the success rate of the manufacturing task with a preset success rate; if the success rate of the manufacturing task is greater than the preset success rate, the accuracy of the energy consumption prediction of the energy consumption optimization model is determined to meet the requirements; if the success rate of the manufacturing task is less than or equal to the preset success rate, the accuracy of the energy consumption prediction of the energy consumption optimization model is determined to not meet the requirements, and the residual feedback gain of the energy consumption prediction is increased.

[0063] Understandably, in the energy consumption optimization control method for embodied robots, a preset success rate is used to characterize the accuracy of energy consumption prediction by the energy consumption optimization model. The core logic is to transform the abstract reliability of energy consumption prediction into a quantifiable success rate range judgment. By comparing the success rate of actual manufacturing task completion with the preset threshold, the accuracy of energy consumption prediction is improved, and the prediction deviation of the model under different task conditions is identified. The preset success rate is set to ensure the stability and practicality of the energy-saving control of the embodied robot. Optionally, the preset success rate is determined through a limited number of experiments by evaluating the effect of different task success rates on the energy-saving control of the embodied robot. The determined preset success rate should be neither too small nor cause excessive interference to the energy-saving control process of the embodied robot. For example, the preset success rate is generally selected in the range of [97.5%, 98.5%].

[0064] Preferably, the preset success rate is 98% in the preferred embodiment.

[0065] Specifically, the success rate of the manufacturing task is the ratio of the number of times the manufacturing task is successfully completed during the execution of the manufacturing task to the total number of manufacturing operations.

[0066] Specifically, successfully completing a manufacturing task means completing all the steps involved in the manufacturing task and ensuring that the product meets the constraints.

[0067] Specifically, the increase in the residual feedback gain of the energy consumption prediction is determined by the difference between the preset success rate and the success rate of the manufacturing task.

[0068] Specifically, when the difference between the preset success rate and the success rate of the manufacturing task is within 1%, the residual feedback gain of energy consumption prediction increases to 1.1 times the original value. When the difference between the preset success rate and the success rate of the manufacturing task exceeds 1%, the residual feedback gain of energy consumption prediction increases by 0.06 for every 1% increase beyond the original value, in addition to the 1.1 times increase. For example, when the difference between the preset success rate and the success rate of the manufacturing task is 3%, the current residual feedback gain of energy consumption prediction is 0.3, and the increased residual feedback gain of energy consumption prediction is 0.3×1.1+0.06×2=0.45.

[0069] In practice, this invention adjusts the residual feedback gain of energy consumption prediction by setting a preset success rate. Since the energy consumption prediction model fails to fully capture the complex energy consumption characteristics during task execution, prediction errors accumulate and affect subsequent control decisions. By increasing the residual feedback gain, the model's response to prediction errors can be enhanced, enabling the energy consumption optimization strategy to quickly correct deviations, improve prediction accuracy, and thus ensure the reliability of the energy consumption optimization strategy execution, further improving the energy-saving control stability of the embodied robot.

[0070] The technical solution of the present invention has been described above with reference to the preferred embodiments shown in the accompanying drawings. However, it will be readily understood by those skilled in the art that the scope of protection of the present invention is obviously not limited to these specific embodiments. Without departing from the principles of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will all fall within the scope of protection of the present invention.

Claims

1. A method for optimizing energy consumption control of a green intelligent manufacturing embodied robot, characterized in that, include: The working condition status data of the embodied robot during the execution of manufacturing tasks is collected, the working condition status data is preprocessed to obtain state features, and the process behaviors in the manufacturing tasks are combined with the state features for mapping and alignment to obtain the process energy consumption feature set corresponding to the process behaviors. The process involves obtaining the constraints of the manufacturing task, training an initial model based on the process energy consumption feature set and the constraints to obtain an energy consumption optimization model, and generating an energy consumption optimization strategy based on the energy consumption optimization model to optimize the energy consumption of the embodied robot. The process also involves obtaining the standard deviation of the periodic energy consumption of the manufacturing task and determining whether the energy-saving control stability of the embodied robot meets the requirements based on the standard deviation of the periodic energy consumption of the manufacturing task. If the energy-saving control stability of the embodied robot does not meet the requirements, the process involves obtaining the fluctuation rate of the motor operating current of the embodied robot to determine whether the robustness of the control strategy meets the requirements. If the robustness of the control strategy does not meet the requirements, the process involves determining whether to reduce the energy consumption constraint disturbance tolerance coefficient of the energy consumption optimization model based on the fluctuation rate of the motor operating current of the embodied robot. If it is not necessary to reduce the energy consumption constraint disturbance tolerance coefficient of the energy consumption optimization model, the process involves determining whether to increase the residual feedback gain of energy consumption prediction based on the success rate of the manufacturing task.

2. The energy consumption optimization and control method for green intelligent manufacturing embodied robots according to claim 1, characterized in that, Determining whether the energy-saving control stability of the embodied robot meets the requirements based on the standard deviation of the cycle energy consumption of the manufacturing task includes: comparing the standard deviation of the cycle energy consumption of the manufacturing task with a preset standard deviation of energy consumption; if the standard deviation of the cycle energy consumption of the manufacturing task is less than or equal to the preset standard deviation of energy consumption, then the energy-saving control stability of the embodied robot meets the requirements; if the standard deviation of the cycle energy consumption of the manufacturing task is greater than the preset standard deviation of energy consumption, then the energy-saving control stability of the embodied robot does not meet the requirements.

3. The energy consumption optimization and control method for green intelligent manufacturing embodied robots according to claim 2, characterized in that, Given that the energy-saving control stability of the avatar robot does not meet the requirements, the robustness of the control strategy is determined based on the fluctuation rate of the motor operating current of the avatar robot.

4. The energy consumption optimization and control method for green intelligent manufacturing embodied robots according to claim 3, characterized in that, Determining whether the robustness of the control strategy meets the requirements based on the fluctuation rate of the motor operating current of the embodied robot includes: comparing the fluctuation rate of the motor operating current of the embodied robot with a preset first fluctuation rate; if the fluctuation rate of the motor operating current of the embodied robot is less than or equal to the preset first fluctuation rate, then the robustness of the control strategy meets the requirements; if the fluctuation rate of the motor operating current of the embodied robot is greater than the preset first fluctuation rate, then the robustness of the control strategy does not meet the requirements.

5. The energy consumption optimization and control method for green intelligent manufacturing embodied robots according to claim 4, characterized in that, Determining whether to reduce the energy consumption constraint disturbance tolerance coefficient of the energy consumption optimization model includes: comparing the fluctuation rate of the motor operating current of the android with the preset first fluctuation rate and the preset second fluctuation rate respectively; if the fluctuation rate of the motor operating current of the android is greater than the preset first fluctuation rate and less than or equal to the preset second fluctuation rate, reducing the energy consumption constraint disturbance tolerance coefficient of the energy consumption optimization model; if the fluctuation rate of the motor operating current of the android is greater than the preset second fluctuation rate, determining that it is not necessary to reduce the energy consumption constraint disturbance tolerance coefficient of the energy consumption optimization model.

6. The energy consumption optimization and control method for green intelligent manufacturing embodied robots according to claim 5, characterized in that, The reduction in the energy consumption constraint disturbance tolerance coefficient of the energy consumption optimization model is determined by the difference between the motor operating current fluctuation rate of the embodied robot and the preset first fluctuation rate.

7. The energy consumption optimization and control method for green intelligent manufacturing embodied robots according to claim 6, characterized in that, Given that the fluctuation rate of the motor operating current of the embodied robot is greater than the preset second fluctuation rate, it is initially determined that the accuracy of the energy consumption optimization model in predicting energy consumption does not meet the requirements, and the accuracy of the energy consumption prediction of the energy consumption optimization model is determined based on the success rate of the manufacturing task.

8. The energy consumption optimization and control method for green intelligent manufacturing embodied robots according to claim 7, characterized in that, The accuracy of the energy consumption prediction of the energy consumption optimization model is determined based on the success rate of the manufacturing task, including: comparing the success rate of the manufacturing task with a preset success rate; if the success rate of the manufacturing task is greater than the preset success rate, the accuracy of the energy consumption prediction of the energy consumption optimization model is determined to meet the requirements; if the success rate of the manufacturing task is less than or equal to the preset success rate, the accuracy of the energy consumption prediction of the energy consumption optimization model is determined to not meet the requirements, and the residual feedback gain of the energy consumption prediction is increased.

9. The energy consumption optimization and control method for green intelligent manufacturing embodied robots according to claim 8, characterized in that, The success rate of a manufacturing task is the ratio of the number of times a manufacturing task is successfully completed during the execution of the manufacturing task to the total number of manufacturing tasks.

10. The energy consumption optimization and control method for green intelligent manufacturing embodied robots according to claim 9, characterized in that, The increase in the residual feedback gain of the energy consumption prediction is determined by the difference between the preset success rate and the success rate of the manufacturing task.

Images